Darwin and Heredity: die Evolution of His
Hypothesis of Pangenesis
GERALD L. GEISON
ITH the publication in 1859 of the Origin of Species,
Charles Darwin could feel quite satisfied that he had
found in natural selection a mechanism to explain how
favored variations are preserved in the struggle for existence. But the problem of just how those variations were
produced in the first place remained elusive as ever. And
variation was by no means the only problem. Acquired characters, embryonic development, the effects of use and disuse, hybridism, reproduction,
reversion—all the phenomena related to heredity were as yet ill-understood,
and a consistent explanatory mechanism was sorely needed.
Darwin had long been aware of the need, and on 27 May 1865 he sent
T. H. Huxley thirty pages of manuscript under the heading, 'The Hypothesis of Pangenesis.'1 Pangenesis was Darwin's attempt to supply the missing
hereditary mechanism, and he was anxious to have the opinion of his good
friend Huxley. Apparently the letter in which Huxley first reported his
opinion of Pangenesis has not survived, but it must have been unfavorable
enough to make Darwin think that he had intended to discourage publication, for on I2july 1865 Darwin wrote him: 'I do not doubt your judgment
is perfectly just and I will try to persuade myself not to publish. The whole
affair is much too speculative... .' 2 But to this Huxley quickly replied that
he had not intended his remarks to prevent or discourage publication. He
This work was carried out with the support of a Traineeship from the U. S. Public Health Service and
a Graduate Fellowship from the National Science Foundation. I am deeply indebted to Dr. Leonard G.
Wilson for numerous suggestions and constant encouragement and to Dr. H. Lewis McKinney for
permission to use the results of his discovery and study of Darwin's annotated copy of Spencer's
Principles of biology (see n. 114 below).
1. Francis Darwin, ed., The life and letters of Charles Darwin (New York, 1899), 2 vols.,n, 227-228.
Hereafter cited as Life and letters. The original letter does not indicate in what year it was sent. Francis
Darwin added in brackets the date [1865?], and this choice is confirmed by Huxley's dated reply (n. 3
below).
2. Ibid., n, 228.
[375]
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
I. INTRODUCTION
376
Journal of the History of Medicine : October 1969
suggested only that Darwin publish his views, 'not so much in the shape of
formed conclusions, as of hypothetical developments of the only clue at
present accessible. . . .' 3
It was in this spirit that Darwin did eventually publish. 'The Provisional
Hypothesis of Pangenesis' became chapter xxvii in The Variation ofAnimals
and Plants under Domestication, the first edition of which is dated 1868.4
Darwin opens that chapter with these words:
It seemed to him that these facts could be connected if one were willing to
accept certain basic assumptions. The first and chief of these assumptions
was that each cell or unit ofan organism throws off minute material particles
—'gemmules' Darwin called them—during each stage of its development.
Darwin supposed that these gemmules 'circulate freely throughout the system, and when supplied with proper nutriment multiply by self-division,
subsequently becoming developed into cells like those from which they
were derived.' He then attributed to the gemmules—in their 'dormant
state'—a 'mutual affinity' by which they were aggregated into the sexual
elements. According to this supposition, the sexual elements must be looked
upon as nothing but a collection of gemmules derived from somatic units.
It is not the reproductive organs which generate new individuals, but rather
the independent units of which an organism is composed. Darwin also
assumed that in the process ofreproduction, the gemmules were transmitted
by means of the sexual elements from parent to offspring. Once transmitted,
they usually developed immediately in the next generation, though not
infrequently they were passed on in a dormant state to future generations,
in whom their development could result in reversion to an ancestral character. In either case, 'their development is supposed to depend on their
union with other partially developed cells or gemmules which precede
them in the regular course of growth.' 6
These were purely hypothetical assumptions, but they lay at the base of
3. Leonard Huxley, ed., Life and letters of Thomas Henry Huxley (London, 1900), 2 vols., 1, 268.
4. Charles Darwin, The variation of animals and plants under domestication (London, 1868), 2 vols.
Hereafter cited as Variation.
5. Ibid., n, 357.
6. Ibid., n, 374.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
In the previous chapters large classes of facts, such as those bearing on budvariation, the various forms of inheritance, the causes and laws of variation, have
been discussed, and it is obvious that these subjects, as well as the several modes of
reproduction, stand in some sort of relation to each other. I have been led, or
rather forced, to form a view which to a certain extent connects these facts by a
tangible method.5
Geison : Darwin and Pangenesis
377
Firstly, on the deficiency, superabundance, fusion, and transposition of gemmules,
and on the redevelopment of those which have long been dormant. In these cases
the gemmules themselves have undergone no modification.... Secondly, in the
cases in which the organisation has been modified by changed conditions, the increased use or disuse of parts, or any other cause, the gemmules cast off from the
modified units of the body will be themselves modified, and, when sufficiently
multiplied, will be developed into new and changed structures.8
One group of causes was basically quantitative—with unmodified gemmules being rearranged and redistributed—and the second group of causes
was qualitative, with the gemmules themselves undergoing substantial alterations in their attributes corresponding to alterations in the conditions of
life. It was the latter group of causes which allowed Darwin to incorporate
into Pangenesis his belief that acquired characters are heritable.
The fact that both parents contribute gemmules accounts for blending
inheritance. But what about those notable exceptions to the doctrine of
blending inheritance—what about prepotency and what about reversion?
Prepotency, says Darwin, results from the gemmules of one parent having
some advantage in 'number, affinity, or vigor' over those of the other
parent.9 Reversion—as well as the appearance of vestigial organs and the
occasional emergence of latent sexual characteristics—can be attributed to
the development of gemmules which had merely lain dormant.10
The gemmules are also brought to bear on the development of bodily
structures in the offspring, normal development depending upon the development ofgemmules in the proper order. Organs abnormally multiplied or
transposed are the result of gemmules being developed in the wrong place
7. Ibid., n, 357.
8. Ibid., n, 396-397.
9. Ibid., n, 386.
10. Ibid., 11, 397-402. Cf. Sir Gavin de Beer, Charles Darwin: evolution by natural selection (Garden
City, N.Y., 1964), p. 204.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
Darwin's attempt to produce plausible explanations of the multitude of
hereditary phenomena 'at present left disconnected by any efficient cause.'7
As the proximate agents of this attempt, the gemmules were put to work in
versatile and ingenious fashion.
Since for Darwin the gemmules eventually develop into cells like those
from which they came, they are first of all made responsible for the resemblances observed between parent and offspring. But, at the same time, they
are also made responsible for the variations which arise. For Darwin, variability depended:
378
Journal of the History of Medicine : October 1969
11. Variation, n, 392.
12. Ibid., n, 398.
13. Ibid.,n, 385.
14. Ibid., u, 38S-387.
15. Ibid., n, 383-384.
16. Francis Galton, 'Experiments in pangenesis, by breeding from rabbits of a pure variety, into
whose circulation blood taken from other varieties had previously been largely transfused,' read 30
March 1871, Proc. toy. Soc. 1872, 19, 393-410.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
in the offspring, of their 'uniting with wrong cells... during their nascent
state.'11 Mutilations are rarely inherited because there is built up, in the
course of time, an excess or reserve of gemmules sufficient to reproduce
even a part which has been repeatedly amputated. Those mutilations which
are inherited arise generally, perhaps exclusively, in connection with disease. Perhaps this is because the gemmules of the lost part are attracted to the
diseased surface and are destroyed there to such an extent that there are not
enough gemmules left for the structure to be developed in the offspring.12
Also explained by gemmule action are regeneration and the repair of
injuries. Both are served by a reserve of gemmules diffused throughout the
tissues of every animal.13
Pangenesis,' wrote Darwin, 'does not throw much light on Hybridism,
but agrees well with most of the ascertained facts.' Because the gemmules of
distinct species have an affinity for their own kind, crossed forms often exhibit unblended characters in 'stripes or blotches.' The generally observed
sterility of the hybrid offspring from two distinct species is not surprising in
view of the delicate nature of the affinities involved. It depends on the
reproductive organs being specially affected in such a way that the hybridised cell-gemmules 'fail to become aggregated within the reproductive
organs.'14
And the 'antagonism which has long been observed . . . between active
growth and the power of sexual reproduction... is partly explained by the
gemmules not existing in sufficient numbers for both processes.' The various forms of reproduction 'graduate into each other and agree in their
product' because they all depend on the aggregation of gemmules derived
from the whole body.15
Such, in brief review, were the explanations which Darwin offered for
the more important hereditary phenomena. Many ofhis explanations sound
remote and irrelevant today. 'Mendelian' genetics offers a simpler and more
persuasive answer to the problem; and placed beside his lasting principle of
natural selection, Darwin's short-lived hypothesis of Pangenesis seems to
pale into insignificance. Notable among the early attempts to test Pangenesis experimentally were the blood transfusion experiments of Darwin's
cousin, Francis Galton.16 By transfusing blood (and therefore gemmules)
Geison : Darwin and Pangenesis
379
17. Recent examples include Gertrude Wichler, Charles Darwin, the founder of the theory of evolution
and natural selection (Oxford, N.Y., 1961); Gertrude Himmelfarb, Darwin and the Darwinian revolution
(Garden City, N.Y., 1959); and Ruth Moore, Charles Darwin, a great life in brief (New York, 1955).
18. See, e.g., R. C. Olby, Origins ofMendelism (London, 1966), pp. 69, 99; Loren Eiseley, Darwin's
century; evolution and the men who discovered it (Garden City, N.Y., 1959), p. 217; Sir Arthur Keith,
Darwin revalued (London, 1955), p. 170; and C. D. Darlington, 'Purpose and particles in the study of
heredity,' in Science, medicine, and history: essays on the evolution of scientific thought and medical
practice...
in honour of Charles Singer, collected and edited by E. A. Underwood (London, 1953), 2 vols., n,
472-481, on p. 474.
19. See esp. C. D. Darlington, Darwin's place in history (New York, 1961). Darlington is especially
critical of Darwin because (pp. 38-39) he 'fell back on the assumption that acquired characters were
inherited, [and] set to work to invent a mechanism for Lamarckian inheritance.'
20. Olby (n. 18), p. 61, asserts that the view was widely held among Darwin's contemporaries that
acquired characters are not inherited, but de Beer (n. 10), p. 204, argues that Darwin 'had no reason to
doubt' the reality of the phenomenon. Perhaps it is safest to say that opinion was divided among
influential nineteenth-century authors: James Cowles Pritchard, William Lawrence, and Joseph
Hooker appear to have denied the inheritance of acquired characters; but (besides Darwin) Charles
Lyell, Robert Chambers, and Herbert Spencer seem to have accepted the phenomenon as real. [See
Conway Zirkle, "The early history of the idea of the inheritance of acquired characters and of pangenesis,' Trans. Amer. phil. Soc, 1946, n. s. 35, 91-151, on pp. 116-119.] Moreover, Darwin offered
(with considerable caution) extensive evidence for his belief, even though much of it concerned rare
phenomena and was accepted rather uncritically.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
from one breed ofrabbit to another, Galton hoped to induce variation in the
offspring of the second breed. These experiments, like all that followed,
failed utterly to confirm Darwin's view; and when, in addition, the idea of
the inheritance of acquired characters became discredited, Pangenesis was
rapidly swept out of court by more satisfying explanations (in particular
Weismann's theory of the continuity of the germ plasm).
As a result, Pangenesis has often been looked upon as one of those mysterious and inexplicable failures of genius. Perhaps because they wish to
present only Darwin's genius, several of his biographers fail to mention
Pangenesis at all.17 Even when it has received some attention, Pangenesis
has generally been maligned as an ad hoc 'Lamarckian' device framed solely
to account for the now-defunct notion of the inheritance of acquired characters.18 Interpreted in this way, Pangenesis seems to corroborate the increasingly Lamarckian drift of Darwin's thought in later editions of the
Origin, and therefore serves those debunkers who wish to assign Darwin a
new and lower place in history.19
But the history of the development of Darwin's ideas on Pangenesis
reveals that this argument focuses too much attention on one aspect of his
hypothesis. It is true that Darwin specifically allowed for the inheritance of
acquired characters by his assumption that environmentally modified cells
give off correspondingly modified gemmules. Whether this constitutes a
real failure on Darwin's part is debatable,20 but it is certain—and more
important—that the inheritance of acquired characters was only one of
380
Journal of the History of Medicine : October 1969
many phenomena which Pangenesis was designed to explain. A variety of
considerations and influences were involved in the development of Darwin's
hypothesis. Only by examining in detail the chronological development of
Darwin's ideas can we begin to assign these influences and considerations
their proper role in the story.
II. THEORIES ON THE ORIGIN OF D A R W I N ' S PANGENESIS
Fleeming Jenkin s 1867 review of the Origin
West's suggestion, never very popular, was succeeded by the view that
Pangenesis was a hastily contrived ad hoc hypothesis designed to answer
certain attacks on the principle of evolution by natural selection. The
spokesmen for this school of thought made special reference to the famous
review of the Origin in June 1867 by Fleeming Jenkin.24 Since this attack
took place shortly before publication of the Variation, and since it was directed especially against the consequences of the hereditary ideas of the
Origin, some writers assumed that Jenkin must have been the inspiration for
21. Geoffrey West, Charles Darwin, a portrait (New Haven, 1938), p. 272.
22. Life and letters, n, 255.
23. Ibid., n, 262-263.
24. Fleeming Jenkin, "The origin of species,' N. Brit. Rev., 1867,4 6,277-318- The review is unsigned
but was later attributed to Jenkin. See W. E. Houghton, ed., The Wellesley index to Victorian periodicals
(Toronto, 1966), p. 691, item 842. Cf. Fleeming Jenkin, Papers literary, scientific, etc.ed. Sidney Colvin
and J. A. Ewing, with a memoir by Robert Louis Stevenson (London, 1887), 2 vols., 1, 215-263.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
Thirty years ago, Darwin's biographer Geoffrey West suggested that Pangenesis was 'conceived as early as 1840 or 1841, [and was] long treasured
and reflected upon in secret. . . .' 21 This view had at least one impressive
piece of evidence; namely, a letter to Charles Lyell written in 1867 in which
Darwin says that Pangenesis is '26 or 27 years old.'22 However indisputable
this sounds atfirst,Darwin must have meant that he had been trying for 26 or
27 years to formulate some hereditary mechanism. It is highly improbable
that he could have achieved as early as 1840 any hypothesis really comparable to Pangenesis as finally presented. This interpretation is supported by
at least two ofDarwin's letters. In February 1868 he wrote Hooker, 'perhaps
I feel the relief [from the hypothesis of Pangenesis] extra strongly from
having during many years vainly attempted to form some hypothesis.' The
day before, he had written almost identically to Wallace, 'It has certainly
been an immense relief to my mind; for I have been stumbling over the
subject for years, dimly seeing that some relation existed between the various classes of facts.'23
Geison : Darwin and Pangenesis
381
Kelvin's 'Age of the Earth' argument
One of the important arguments which Jenkin brought forth in 1867
against natural selection antedated his review by several years. This was the
25. Eiseley (n. 18), p. 217.1 have omitted from Eiseley's passage the name 'Bennett.' But it is perfectly obvious, from Eiseley's own account, that A. W. Bennett's critique of natural selection had
nothing whatever to do with the original development of Pangenesis. It did not appear until 1870, two
years after the Variation was published. See Eiseley's footnote 11, p. 215.
26. Peter Vorzimmer, 'Charles Darwin and blending inheritance,' his, 1963,34, 371-390, on p. 386.
27. R. C. Olby, 'Charles Darwin's manuscript of Pangenesis,' Brit.J. Hist. Sci., 1963, 1, 251-263,
on p. 251.
28. Vorzimmer's point is valid even though he incorrectly states that the Variation was sent to the
printers on 21 November 1866. That is, rather, the date on which Darwin reported that he had
finished Pangenesis. Actually, all of the chapters of the Variation except the last were sent to the printers on 21 December 1866. See Sir Gavin de Beer, ed., 'Darwin's Journal,' Bull. Brit. Mus. (nat. Hist.),
1959, hist. ser. 2, 1-21, on p. 17.
29. Vorzimmer further claims (n. 26), p. 386, that Darwin, 'having been specifically instructed by
the printer not to make any but the most minor alterations,... did indeed make very few changes.'
His documentation for this claim is improper. The published evidence seems to indicate nothing more
than that Darwin received the first proofs of the Variation on 1 Marchi867and finished correcting all
the proofs on 15 November 1867. See de Beer (n. 28), p. 17. Jenkin's review appeared in June 1867.
30. Olby (n. 27), pp. 253-263.
31. Both in the 1865 manuscript and in the 1868 version, the 'chief points are the modes of reproduction, variation, inheritance, reversion, and the effects of use and disuse.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
Darwin's new hereditary mechanism. Loren Eiseley, for example, charged
in 1959 that Darwin's elaboration of Pangenesis was 'an indirect escape
from such problems a s . . . Jenkin had formulated.'25
In 1963 both Peter Vorzimmer26 and R. C. Olby 27 argued against this
point of view, pointing out that Darwin had already sent the manuscript of
Pangenesis to Huxley in 1865, two years before Jenkin's review appeared.
Vorzimmer also pointed out that by the timeJenkin's review was published,
Darwin had already sent the Variation to the printers.28 Thus, if anything
that appears in the first edition of the Variation bears Jenkin's influence, it
could only have been inserted by Darwin after he received the proofs from
the printer.29
This evidence alone makes it highly improbable that Jenkin's review
exerted any significant influence on Darwin's hereditary hypothesis. The
case is clinched by Olby's publication of Darwin's 1865 manuscript of
Pangenesis.30 For despite certain differences, especially in size, the hypothesis which Darwin called Pangenesis was in its 1865 version strikingly similar
to the version published in 1868, both with respect to its basic assumptions
and to the chieffacts it was designed to connect by some 'intelligible bond.'31
Jenkin's review of the Origin could hardly have been the inspiration for a
hypothesis which was fully developed at least two years before his review
appeared.
382
Journal of the History of Medicine : October ig6g
32. Seejenkin (n. 24), pp. 294-305.
33. Ibid., p. 301. See also n. 42 below.
34. See Eiseley (n. 18), pp. 233-241, esp. 240.
35. See S. P. Thompson, The life of William Thomson, Baron Kelvin ofLargs (London, 1910), 2 vols.,
I, 41-42, 185-188, 535-53<>36. William Thomson, 'Physical considerations regarding the possible age of the sun's heat,' Rep.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
argument based on William Thomson's estimates ofthe age ofthe earth and
of the age of the sun's heat.32 On strictly physical considerations, Thomson
(later Lord Kelvin) had been arguing for several years that the uniformitarian geologists (including Lyell and Darwin) were not justified in their
assumption that geology had at its disposal 'almost unlimited time.' Depending on which physical assumptions were adopted for the purposes of
calculation, Thomson's estimates varied greatly; but all estimates ledjenkin
to the conclusion that, 'Not only is the time limited, but it is limited to
periods utterly inadequate for the production of species according to Darwin's views.'33
It has been suggested, by Eiseley for example, that Kelvin's 'age of the
earth' argument against natural selection forced Darwin to consider the
possibility that variation took place (or had taken place) more rapidly than
natural selection alone could explain; and that, in the face of this objection,
Darwin retreated in later editions of the Origin to more Lamarckian hereditary views, including the concept ofthe inheritance ofacquired characters.34
Although Eiseley does not specifically suggest that the hypothesis of Pangenesis was originally designed to meet Kelvin's argument, this possibility
should not be ignored, especially since Kelvin's possible influence cannot be
rejected out of hand, as can Jenkin's, on the basis of chronology. Almost
four full years before Darwin sent his manuscript of Pangenesis to Huxley,
Kelvin had begun to argue explicitly and concretely against the assumptions
that the uniformitarian geologists were making about the age of the earth
and its sun.
The story of Kelvin's attack on uniformitarian geology begins in earnest
at the 1861 meeting of the British Association, even though Kelvin and his
biographer Silvanus Thompson were able in retrospect to find vague traces
of his argument in five previous publications, beginning as early as 1842,
and in an unpublished inaugural dissertation of 1846.35 At that meeting in
1861, Kelvin argued that the sun was an incandescent molten mass whose
energy was dissipating at a rapid rate. Invoking the second law of thermodynamics, he estimated that it was 'on the whole most probable that the sun
has not illuminated the earth for 100,000,000 years, and almost certain that
he has not done so for 500,000,000 years.'36
Geison : Darwin and Pangenesis
383
Brit. Ass., 1861 [imprint 1862], trans, of the sections, pp. 27-28, quote on p. 28. This was only an
abstract of Thomson's paper, but the full text appears in Thomson, 'On the age of the sun's heat,'
Macmillan's Mag., 1862, 5, 388-393, quote on p. 393. Cf. Thomson (n. 35), i, 421. The reference in
note 39 below is to this unabstracted version of Thomson's paper.
37. William Thomson, 'On the secular cooling of the earth,' read 28 April 1862, Trans, roy. Soc.
Edinb., 1864, 23, 157-169; cf. Phil. Mag., 1863, 25, 1-14. This argument was later repeated, in arrogantly short form, in Thomson, "The "doctrine of uniformity" in geology briefly refuted,' read 18
December 1865, Proc. roy. Soc. Edinb., 1866, 5, 512-513.
38. See Thomson (n. 37), p. 159.
39. See Thomson (n. 36), pp. 391-392.
40. William Thomson, 'On geological time,' read 27 February 1868, Trans, geol. Soc. Clasg., 1871,
5, 1-28.
41. See, e.g., Archibald Geikie, 'Address to the geological section,' Rep. Brit. Ass., 1899 [imprint
1900], pp. 718-730, on p. 722.
42. For a partisan review of the controversy as it stood in July 1869, see [P. G. Tait), 'Geological
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
In April of 1862 Kelvin read a paper at the Royal Society ofEdinburgh in
which he set forth a second major objection to those geologists who believed that the earth had existed for almost infinite stretches of time. In this
paper he turned from the sun to the earth and argued on the basis of calculations of heat loss from the earth's crust that uniformitarian geologists could
not claim the vast amount of time required by their views. He estimated
that the earth had existed as a consolidated planet for only 20 to 400 million
years.37
Kelvin published nothing further on the question of geological time
until after Darwin had already sent Huxley his manuscript of Pangenesis.
But these two papers already contain the two most influential arguments
which guided all of Kelvin's later attacks against uniformitarian geology
and evolutionary biology. Moreover, because the second paper mentions
Lyell by name,38 while the first specifically criticizes Darwin's estimate of
300,000,000 years for the 'denudation of the Weald,'39 Darwin's attention
might well have been drawn to them. But is there any real evidence that
Pangenesis was designed to answer Kelvin's objections?
The answer is no, for there exists at present no published evidence that
Darwin took any notice whatever of Kelvin's objections before 1868. As a
matter of fact, Kelvin's attack on uniformitarian geology apparently impressed almost no one (except Fleeming Jenkin, who was his student) until
after he delivered a major address before the Geological Society of Glasgow
in February 1868. In this address Kelvin for the first time pulled together
his two earlier arguments against the doctrine of uniformity and presented a
new, third argument, based on calculations of the tidal retardation of the
earth's rotation.40 This address apparently caused great concern among
uniformitarian geologists41 and initiated a lively controversy.42 It provoked
384
Journal of the History of Medicine : October ig6g
Robert Olby's recent theory
Olby's interesting and original interpretation of the origin of Darwin's
Pangenesis is based to an important extent on his analysis ofDarwin's manuscript of 1865. In his book The Origins ofMendelism (1966), Olby states his
position to be that 'the hypothesis represents the crystallization of Darwin's
thoughts over a period ofa quarter ofa century, thoughts which began with
his wonder at the ability of a planarian to regenerate after division.'47 He
believes that Darwin began with 'the assumption of the identity of all forms
time,' N. Brit. Rev., 1869, 50, 406-439. The review is unsigned but is attributed to Tait in Houghton
(n. 24), p. 693, item 910. Among the participants in the dispute was Joseph Hooker, who defended the
uniformitarian view in his presidential address to the British Association in 1868 (August). See Rep.
Brit. Ass., 1868 [imprint 1869], pp. lviii-lxxv, on lxx-lxxii. Hooker does not respond directly to
Kelvin's famous address of the previous February but rather to Kelvin's position as it is presented by
Jenkin in his 1867 review of the Origin. Because we now emphasize Jenkin's 'swamping effect' argument against natural selection, it is interesting that Hooker refers instead to the 'age of the earth'
argument as 'the most formidable . . . urged by the reviewer,' (lxxii) though he dismisses it with
apparent confidence.
43. T. H. Huxley, "The anniversary address of the president,' Quart.], geol. Soc. Lond., 1869, 25,
xxviii-liii, on xxxviii-liii.
44. William Thomson, 'Of geological dynamics,' read 5 April 1869, Trans, geol. Soc. Clasg., 1871,
3, 215-240, quotation on p. 222.
45. See Life and Letters, n, 296, 326; Francis Darwin and A. C. Seward, eds., More letters of Charles
Darwin (London, 19035,2 vols., 1,313-314,11,6-7,163-164,211-212. [Hereafter cited as More letters.]
Cf. Eiseley (n. 18), pp. 235, 237, 240, 242.
46. See More letters, 1, 314; Morse Peckham, ed., The Origin of species by Charles Darwin: a variorum
text (Philadelphia, 1959), pp. 486, 513. In the sixth and last edition, Darwin referred to Kelvin's
objection as 'probably one of the gravest as yet advanced.' Ibid., p. 728.
47. Olby (n. 18), p. 100.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
Huxley to a detailed reply,43 which in turn inspired a rejoinder from Kelvin.
It was in this rejoinder that Kelvin first converted his objection to uniformitarian geology into an explicit objection against evolution by natural selection. 'The limitations of geological periods, imposed by physical science,'
he wrote, 'cannot, of course, disprove the hypothesis of transmutation of
species; but it does seem sufficient to disprove that transmutation has taken
place through "descent with modification by natural selection." >44
From 1868 on, Darwin's letters reveal his increasing disquietude at Kelvin's objections,45 and even as early as 1869 he was so impressed by them
that he faced the problem openly in the fifth edition of the Origin.46 But
however much Darwin may have been influenced by Kelvin's objections
after 1868, there is no evidence that he had been influenced by them before
that date. There is, therefore, no reason to believe that Pangenesis had its
origin in Kelvin's age of the earth argument, any more than it did injenkin's
review.
Geison : Darwin and Pangenesis
385
48
III.
THE EVOLUTION OF PANGENESIS
The period of early concern: 1837-18^
In July 1837 Darwin opened his first notebook on the transmutation of
species. By the time he closed the fourth and last of these notebooks twentyfour months later, he had formulated the basic elements of his theory of
evolution by natural selection.50 But natural selection needs a supply of
variations on which to work, and the problem of how he might account for
these variations naturally occurred to Darwin at the same time that his ideas
on natural selection were beginning to develop. Among the problems
which he considered in hisfirstnotebook were reproduction, variation and
its causes, hybridism, prepotency in crosses—virtually all of the chief problems of variation and heredity, with the exception of sterility and the inheritance of the effects of use and disuse.51 In the later notebooks, these
problems, too, attracted Darwin's attention.
The information then available on variation and heredity was fragmentary and often contradictory; the causes of variation were wholly unknown. In his search for enlightenment, Darwin went beyond the existing
literature on the subject; in correspondence and in conversation, he asked
breeders and gardeners to convey to him the results of their experience. To
expedite matters, he eventually composed an eight-page quarto pamphlet,
48. Ibid., p. 99.
49. Ibid., pp. 87, 93-95, quote on p. 95.
50. See Sir Gavin de Beer, ed., 'Darwin's notebooks on transmutation of species,' in 4 parts, Bull.
Brit. Mus. (pat. Hist.), i960, hist. ser. 2, 23-183.
51. Ibid., p. 39.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
ofreproduction,' and proceeded inductively to Pangenesis. He documents
this assertion in chapter four of his work, suggesting that Darwin needed to
deny the basic distinction between sexual and asexual processes in order to
explain 'how acquired characters can be inherited [from the somatic cells]
by the sex cells and therefore how changes in the conditions of life can give
rise to heritable variation in sexually reproducing organisms.'49
Olby's argument, though suggestive and persuasive as far as it goes,
concentrates too exclusively on one aspect of the story. He does not develop
in detail the steps by which Darwin came to his hypothesis, nor does he
discuss the possible influence on Pangenesis of contemporary theories and
writers, with the notable exception of Charles Naudin. In the end, Olby's
argument suffers from the old tendency to associate Darwin's Pangenesis too
exclusively with 'soft' or Lamarckian heredity. Olby's interpretation is
examined in greater detail in the section on Naudin below.
386
Journal of the History of Medicine : October 1969
'Questions about the Breeding of Animals,' printed privately in May
1839. 52
52. This rare pamphlet has now been twice republished—once in facsimile, edited and with an
introduction by Sir Gavin de Beer, Sherbom Fund Facsimile No. 3, Dec. 1968, Society for the
Bibliography of Natural History; and as an appendix (pp. 277-281) to Peter Vorzimmer, 'Darwin's
Questions about the breeding of animals,'J. Hist. Biol., 1969, 2, 269-281.
53. See Vorzimmer (n. 52), pp. 273-276, quote on p. 275.
54. Ibid., p. 276.
55. Charles Darwin, On the origin of species, a facsimile of the first edition with an introduction by
Ernst Mayr (New York, 1967), pp. 12-13. Hereafter cited as Origin.
56. Ibid.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
From an analysis of this pamphlet and of the transmutation notebooks,
Peter Vorzimmer has recently suggested that by the second half of 1839,
Darwin already had in mind 'a limited hypothesis regarding the nature of
inheritance.'53 But this early 'hypothesis' was very limited indeed: it was at
most a crude suggestion about a very specific problem—namely, how a
new character could be transmitted and persist in the face of blending inheritance.54 It was not a general hypothesis for heredity; it said nothing of
gemmules; it offered no proximate mechanism for inheritance. In short,
this 'limited hypothesis' was not Pangenesis, nor any precursor of it. Neither
the 'Questions' nor the transmutations notebook contain any of the essential ideas of Pangenesis. They reveal only Darwin's early, deep concern
about the nature and causes of hereditary phenomena.
This early concern continued undiminished during the years Darwin was
carefully working out his principle of evolution by natural selection. In 18 59
Darwin devoted a full chapter of the Origin to hybridism. He also devoted
chapters to variation under domestication, to variation under nature, and
to the laws of variation. In the chapter on variation under domestication,
Darwin recognized that insofar as natural selection was concerned, 'any
variation which is not inherited is unimportant....' And emphasizing that
the fundamental operational axiom of breeders was that 'like produces
like,' he suggested that 'perhaps the correct way of viewing the whole
subject, would be, to look at the inheritance of every character whatever as
the rule, and non-inheritance as the anomaly.'55
Darwin was, however, still unable to offer much in the way of theoretical
explanation. Thus in the chapter 'Variation under Domestication' he says:
'The result of the various, quite unknown, or dimly seen laws ofvariation is
infinitely complex and diversified.... The laws governing inheritance are
quite unknown.' 56 Similarly, the chapter, 'The Laws of Variation,' opens
with these words: 'I have hitherto sometimes spoken as if the variations...
had been due to chance. This, of course, is a wholly incorrect expression,
Geison : Darwin and Pangenesis
387
The period of rapid development: 1860-1865
The first edition of the Origin was published on 24 November 1859.
During the last two months of that year, Darwin was preparing a second
edition of the Origin and was otherwise occupied by 'an enormous correspondence.'60 But then in January i860 he began to look over manuscripts and to arrange his notes for work on The Variation of Animals and
Plants under Domestication.61 This was to be an extended and more completely documented version of thefirstchapter of the Origin. At last Darwin
was able to concentrate on arranging and gathering information about
those problems ofvariation and heredity which had been bothering him for
more than twenty years. Obviously he had not included in the Origin all of
his information on variation and when in i860 he began work on his new
book, he already had a solid core of notes. But he did not rest content
with this core; using mostly contemporary sources, he was able to accumulate a vast amount ofnew information—a new basis for attacking the problems which had concerned him ever since he had opened his first transmutation notebook in 1837.
57. Ibid., p. 131.
58. Ibid., pp. 131-170; cf. ibid., p. 43, where Darwin mentions with regard to domestic races the
same factors later discussed more generally and in more detail, and where he adds the observation that
'in some cases, I do not doubt that the intercrossing of [distinct] species . . . has played an important
part in the origin of our domestic productions.'
59. Ibid., p. 167.
60. Nora Barlow, The autobiography of Charles Darwin (1809-1882), with original omissions restored, edited with appendix and notes (New York, 1959), p. 126.
61. See ibid., pp. 126-127. Cf. de Beer (n. 28), p. 15.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
but it serves to acknowledge plainly our ignorance of the cause of each
particularvariation.'57Heascribedsome influence to the external conditions
of life, especially as acting on the reproductive system; more important
seemed to be habit and the effects of use and disuse; he gave the principle
of the 'correlation of growth' some role; and several lesser factors seemed
to be involved.58 But in the end, Darwin was forced to conclude that 'our
ignorance of the laws of variation is profound. Not in one case out of a
hundred can we pretend to assign any reason why this or that part differs,
more or less, from the same part in the parents.'59
Thus Darwin's abiding interest in the problems of variation and heredity
had as late as 1859 produced little in the way of theoretical structure. There
is no indication that he had as yet hit upon the main features of Pangenesis.
Yet by May 1865 he was able to send Huxley a manuscript of his views in
rather fully developed form. What had happened in the intervening years?
388
Journal of the History of Medicine : October ig6g
With respect to reversion, I have been raking up vague recollections of vague
facts; and the impression on my mind is rather more in favour of reversion than it
was when you were here.... I have not as yet put all my facts on this subject in
mass, so can come to no definite conclusions.66
By March 1863 his 'raking up' and his exchange of letters with Hooker on
the question of reversion had brought him to his view of'latency,' which he
now saw could be applied to a number of facts:
This view of latency collates a lot of facts—secondary sexual characters in each
individual; tendency of character to appear temporarily in youth; effect of crossing in educing talent, character, etc. When one thinks of a latent character being
handed down, hidden for a thousand or ten thousand generations, and then suddenly appearing, one is quite bewildered at the host of characters written in invisible ink on the germ.67
During this period, Hooker was arguing for 'an inherent tendency to
vary, wholly independent of the physical conditions,'68 but Darwin attached increasing importance to the direct effects of conditions, and especially to the effects of use and disuse. In a letter written to Hooker in March
1862 Darwin announced his intention to present a host offacts and measurements relating to the effects of use and disuse: 'Whenever my book [the
Variation] on poultry, pigeons, ducks, and rabbits is published, with all the
measurements and weighing of bones, I think you will see that "use and
disuse" at least have some effect.'69
62. See Origin, pp. 159-167.
63. Ibid., esp. p. 473.
64. De Beer (n. 50), p. 39. De Beer includes the inheritance of the effects of use and disuse in a list
of the 'chief subjects missing' from the extant portion of Darwin's first notebook.
65. See Origin, pp. 11, 134-139.
66. More letters, 1, 120.
67. Ibid., I, 473.
68. Ibid., I, 198.
69. Ibid., 1, 199.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
As Darwin accumulated data, his interest in the phenomenon of reversion
and in the inheritance of the effects of use and disuse grew. In the Origin
he had described in some detail cases of reversion to zebra-like stripes
in horses and of reversion to blue color and black markings in pigeons,62 but
his real concern here was not the laws of inheritance, but the evidence that
this reversion offered for the view that horses and especially pigeons had
descended from a single (or very few) primordial stock(s).63 The effects of
use and disuse—totally absent from Darwin's first transmutation notebook—64 managed to receive several pages of attention in the Origin.65
After 1859 we find Darwin attaching increasing importance to both
phenomena. In May 1859 he wrote to Hooker:
Geison : Darwin and Pangenesis
389
By the time Darwin was ready to draft his manuscript of Pangenesis in
May 1865, reversion and the effects of use and disuse had become perhaps
the two most important of the many hereditary phenomena which he
sought to connect 'by some intelligible bond.' In the manuscript then dispatched to Huxley, Darwin indicates the importance which he now attached to reversion:
In the same passage Darwin repeats those examples of reversion in horses
and pigeons which he had given in the Origin and suggests briefly that
reversion also occurs during the growth of individual fowl, cattle, and
rabbits. He mentions, too, the sudden development of vestigial organs believed to be possessed by an early ancestor, 'as with additional mammae in a
woman and a fifth stamen in some Scrophulariaceae.' And, 'in nearly the
same way,' secondary male characteristics can become manifest in females,
an analogy which 'is the more appropriate from the case recorded of a hen,
which assumed some of the masculine characters, not ofher own breed, but
of her wild progenitor.'
By 1865, then, Darwin had added a few specific examples of reversion to
his Origin account and had included under reversion the sudden appearance
of long-lost organs and the occasional appearance of secondary sexual traits
traditionally belonging to the opposite sex. Darwin was now thinking of
reversion as a general hereditary phenomenon, recognizing in it a plausible
explanation for the great majority ofcases ofnon-inheritance.71 Non-inheritance was still for Darwin the anomaly, but the phenomenon of reversion
—brought on by the sudden development of latent gemmules—could now
account for the vast majority of these anomalies. From the Variation it is
clear that these brief remarks of 1865 were based on a much larger catalogue
offacts which he had been carefully gathering for some time, and which had
especially attracted his attention after 1859.72
70. Olby (n. 27), pp. 256-257.
71. Ibid., p. 256.
72. Of the approximately seventy different works cited by Darwin in his chapter on reversion in
the Variation (n, 28-61), more than thirty were published between 1859 and 1865.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
This principle of Reversion is the most wonderful of all the attributes of Inheritance. It frequently comes into action. What can be more wonderful than that
characters, completely lost during scores or hundreds or even thousands of generations, should suddenly reappear perfectly developed
In every living creature we may feel assured that a host of lost characters lie latent and ready to be
evolved under the proper conditions. How can we make intelligible and connect
with other facts this wonderful and common capacity of reversion—this power of
apparently calling back to life long lost characters.70
39°
Journal of the History of Medicine : October 1969
The language of the manuscript of 1865 also reveals Darwin's increasing
wonder at the heritable effects of use and disuse:
In suggesting briefly these examples of the inherited effects of use and disuse
in ducks, horses, rabbits, and dogs, Darwin had repeated only one example
(that of ducks) from the Origin; and though he had at the same time chosen
to ignore several other examples given in the Origin,14 this was probably due
rather to his wish to keep the manuscript short than to any loss of interest in
the phenomena.
Darwin had in fact become more and more impressed with the heritable
effects of use and disuse, especially in domesticated animals. Besides his own
painstaking measurements of these effects on the capacity of skulls in domesticated rabbits, and on the size and weight of limbs and skeletons in
domesticated pigeons, ducks, fowl, and rabbits,75 Darwin included in the
Variation several new examples gleaned from works published between
1859 and 1865: for example, the effects of use and disuse on the legs, snout,
jaws, and teeth ofdomesticated pigs (effects said to be 'strictly inherited'); on
the lungs and livers of domesticated cattle; and on the nature of stomach
coats and the length of intestines due to altered diets in both wild and domestic animals.76 As in the Origin, Darwin's presentation was cautious and
reasonable. He emphasized the difficulty of distinguishing between the
73- Olby (n. 27), pp. 257-258.
74.. See Origin, pp. 11,134-139. Examples of the effects of use and disuse which Darwin had noted
in the Origin but ignored in the manuscript of 1865 include the drooping ears of domesticated animals
and the well-developed udders of domesticated goats and sows (p. 11); the loss of feet in dung-feeding
beetles, the underdeveloped wings of some Madeira beetles, and the underdeveloped or missing eyes
in moles, burrowing rodents, crabs, and cave-dwelling rats and insects (pp. 134-139).
75. See Variation,!, 124-128,171-177,270-274,284-286; cf. iiiU.n, 298, where Darwin summarizes
his conclusions.
76. Ibid., n, 299-303.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
. . . But what shall we say about the inherited results of the use and disuse of particular organs? The domesticated duck flies less and walks more than the wild
duck, and the limb bones of the tame duck have become in a corresponding manner diminished and increased . . . : a horse is trained to certain paces, and the colt
inherits similar consensual movements: the domesticated rabbit becomes tame
from close confinement; the dog intelligent from associating with man; the retriever is taught to fetch and carry, and these mental endowments are inherited.
Nothing in the whole circuit of natural history is more wonderful. How can the
use or disuse of a particular limb or of the brain affect a small aggregate of cells in
the reproductive organs, in such a manner that the being developed from these
organs inherits these newly acquired characters of either one or both parents.73
Geison : Darwin and Pangenesis
391
The cell theory
One important theoretical development was the theory that the nucleated
cell is the basis of the organization of both plant and animal bodies—a theory
generally associated with Schleiden and Schwann and dating from about
the year 1839. The cell theory is, then, about thirty years older than Pangenesis, but even as late as 1868 it was not fully established for animals.77
Nonetheless, it had by that time gained sufficient acceptance78 for Darwin
to use it, especially in the version of Pangenesis which appeared in the
Variation in 1868, though it was also on his mind when he drafted the
manuscript of 1865.79
Part of the importance of the cell theory was that it allowed Darwin to
think small. Once the existence ofmicroscopic cells was granted, the notion
that even smaller material units might be the bearers of heredity seemed also
more plausible. Perhaps only an increase in the optical power of microscopes was necessary to reveal the Darwinian gemmules! Another reason
for the importance of the cell theory was the evidence it provided for the
77. It is because the 'cellular theory is not fully established' that Darwin hesitates to call bis gemmules 'cell-gemmules.' See ibid., n, 374.
78. What dispute there was centered mainly about the question of the origin of cells. Some disputed
the specific name cell, but as Darwin put it, 'every one appears to admit that the body consists of a
multitude of "organic units," each of which possesses its own proper attributes, and is to a certain
extent independent of all others.' See ibid., n, 370-371.
79. Olby (n. 27), p. 252, says that 'Darwin made no reference to the cell theory' in the manuscript
of 1865; nonetheless Darwin used the word cell fourteen times in that manuscript. In the expanded
version of Pangenesis which appears in the Variation, Darwin specifically discusses the cell theory,
referring on several occasions to Claude Bernard's Lecons sur Us proprUUs its tissus vivants (Paris, 1866)
and to the English translation of Rudolf Virchow's Cellular pathology (see n. 83 below).
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
effects ofnatural selection and use and disuse, but the evidence he had given,
much of it new, led him to the conclusion that many heritable variations
must be at least partly caused by the inherited effects of use and disuse.
Since Darwin became especially interested between 1859 and 1865 in
finding an explanation for the 'wonderful' phenomena of reversion and of
the inheritance of the effects of use and disuse, perhaps he was led to his
hypothesis by reflecting on these new interests and on the new information
he had accumulated about variation and heredity. In fact, however, Darwin's eventual exposition of Pangenesis was influenced not only by this
newly acquired factual information, but also by theoretical developments
which had been going on throughout the years of his search, but which
became especially important during the crucial years between 1859 and
1865.
392
Journal of the History of Medicine : October 1969
80. See Variation, n, 368-371. Darwin here particularly enlists the support of Bernard and Virchow
for his elaboration of the functional independence of the bodily parts.
81. Ibid., 11, 377.
82. In the letter which accompanied the manuscript of Pangenesis, Darwin said he would not
ordinarily ask such a 'very great favour' from one so 'hard worked' as Huxley: 'You must refuse if
you are too much overworked.' Life and letters, n, 227-228.
83. Rudolf Virchow, Cellular pathology as based upon physiological and pathological histology; twenty
lectures delivered in . . . 1858, translated from the 2nd edition of the original by Frank Chance, with
notes and numerous emendations (London, i860).
84. See Variation, n, 370.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
functional independence of the bodily parts.80 To Darwin, it seemed but
'one small step' from the realization of this functional independence to the
assumption 'that each cell casts off a free gemmule, which is capable of reproducing a similar cell.'81
One particular version of the cell theory seems especially important for
Pangenesis—namely that version which asserted the doctrine oiomnis cellula
e cellula. The idea that all cells are derived by division from previously
existing cells is perfectly consistent with, and may even have suggested,
Darwin's assumption that the gemmules thrown off by previously existing
cells self-multiply by division and subsequently become developed into
similar cells. It must be admitted that Darwin does not refer to the doctrine
oiomnis cellula e cellula in the manuscript of 1865, but this may reflect his
wish to keep the manuscript as short as possible so that he would not overburden Huxley.82 Cellular Pathology, the book in which Rudolf Virchow
so forcefully put forth the doctrine, appeared in 1858 and was translated
into English in i860.83 It was the English version which Darwin read, and
the fact that it was available as early as i860 suggests that Darwin may well
have read it before drafting his manuscript of 1865. In the expanded version
ofPangenesiswhichappearedin 1868,Darwinrefersspecifically to the doctrine o£ omnis cellula e cellula and to Virchow's book.
In the spirit of objectivity, Darwin reported in 1868 that there was an
alternative school, which 'maintains that cells and tissues of all kinds may be
formed, independently of pre-existing cells, from plastic lymph or blastema.' Since he had 'not especially attended to histology,' Darwin felt it
would be 'presumptuous' to express an opinion on the two opposed doctrines. Nonetheless, the sources which he consulted about the cell theory
generally favored omnis cellula, and it is difficult to escape the feeling that
Darwin best expressed his own opinion about the genesis of cells when
he said t h a t ' . . . the doctrine of omnis cellula is admitted for plants, and is a
widely prevalent belief with respect to animals.'84 Richard Owen, writing
in 1868, went so far as to say that the gemmule hypothesis was 'absolutely
Geison : Darwin and Pangenesis
393
based on "pre-existence"—or'omnis cellula," &c,'85 and this point of view
was corroborated by a reviewer who suggested that Pangenesis 'starts from
the notion that every organized body is composed of cells . . . capable of
reproducing their own sorts... .' 86
The influence of earlier theories
Precursors before 1800
Darwin's list of earlier pangenetic theorists is by no means exhaustive—
not even in the second edition of the Variation, by which time he had added
to the list Hippocrates, Ray, and a contemporary, Mantegazza.89 Conway
85. Richard Owen, OH the anatomy of vertebrates (London, 1866-68), 3 vols., m, 813.
86. 'Variations of animals and plants under domestication,' Student Intellectual Observer, 1868, 3rd
set. 1, 179-188, on 187.
87. Or, more properly, the Hippocratic corpus. A pangenesis-like hypothesis appears in Perigones
(about 400 B.C.).
88. Variation, n, 374-375.
89. See Charles Darwin, The variation of animals and plants under domestication, authorized ed. [from
Darwin's 2nd ed.] (New York, 1899), 2 vols., n, 359, n. 42.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
Darwin was by no means the first to offer a mechanistic, particulate
theory of inheritance. In fact, pangenesis-like theories of heredity can be
traced back at least as far as Hippocrates.87 To what degree was Darwin
influenced by earlier accounts of such theories?
In the chapter on Pangenesis in the first edition of the Variation, Darwin
acknowledged that his hypothesis was not totally original: 'Nearly similar
views have been propounded, as I find, by other authors, more especially
by Mr. Herbert Spencer; but they are here modified and amplified.'88 In an
extensive and important footnote to this passage, Darwin referred at some
length to the previous views of Buffon, Bonnet, and Owen, as well as
Spencer. Especially significant in the passage itself are the words 'as I find,'
for they seem to imply that Darwin had found out about these similar ideas
only after he had already independently reached his conclusions. In keeping
with this interpretation, the Pangenesis manuscript of 1865 contains no
acknowledgement of previous pangenetic theories.
On the other hand, the absence in the 1865 manuscript ofany reference to
previous theories might be interpreted as another omission to keep the
manuscript as short as possible for Huxley's benefit, or less charitably, as a
deliberate attempt to suppress his indebtedness to others. For these reasons,
it seems advisable to look more carefully at the possible influence ofprevious
theorists.
394
Journal of the History of Medicine : October ig6g
90. Zirkle (n. 20).
91. See, e.g., de Beer (n. 10), p. 205. Darlington, who calls Darwin 'the gamekeeper of natural
selection and the poacher of pangenesis,' prefers to speak pejoratively of Darwin's (apparently intentional) 'disregard of historical propriety or historical knowledge.' See Darlington (n. 18), p. 474.
92. In a total of approximately 2000 entries in Catalogue of the library of Charles Darwin now in the
Botany School, Cambridge, compiled by H. W . Rutherford (Cambridge, 1908), I count forty-three
which are dated prior to the year 1800.
93. Life and letters, n, 26s
94. Ibid., n, 228.
95. Ibid., n, 228-229.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
Zirkle has found more than ninety 'precursors' whose ideas, in one way or
another, bear some resemblance to Pangenesis.90 Indeed Pangenesis, at its
simplest level, had always been the natural and traditional way to account
for blending inheritance and for the inheritance of acquired characters.
But did any of the precursors actually influence Darwin's ideas? The
answer is greatly simplified by the fact that Darwin seems to have had little
historical sense.91 He was certainly not much inclined to read the prenineteenth century literature,92 and there is little chance that he was influenced by pangenetic theories proposed before 1800.
Darwin did not know ofHippocrates' somewhat similar views until 1868
when informed of them by William Ogle. 93 And he apparently had not
read the relevant passages in Buffon and Bonnet until Huxley called them to
his attention after reading the Pangenesis manuscript of 1865. In the letter of
12 July 1865 Darwin thanked Huxley for letting him know about Buffon
and Bonnet, adding: 'It would have annoyed me extremely to have republished Buffon's views, which I did not know of, but I will get the book; and
if I have strength I will also read Bonnet.'94
After reading Buffon, Darwin admitted that 'whole pages are laughably
like mine,' 95 but he was also able to point out that neither Buffon nor
Bonnet had preceded him in his assumption that each unit of living tissue
produced a little gemmule or bud at each stage in its development. Buffon
assumed that his indestructible 'organic molecules,' in many respects similar
to Darwin's gemmules, were animated particles which could be found
distributed throughout nature. Alreadyformed, they needed only to be imbibed with the food into the sap or blood of organisms in order to fulfill
their primary task of being assimilated into the organs which were their
respective analogues. Excess organic molecules, not needed for nourishment,
collected to form buds and the sexual elements. Bonnet used the concept of
somatic 'germs' to account for his observations of regeneration, especially
in earthworms. But in order to account for parthenogenesis in aphids,
Bonnet had postulated the existence of pre-formed germs within germs
(emboitement), 'ready for all succeeding generations.' Darwin, while admitting the similarities, pointed to the fundamental difference between the
Geison : Darwin and Pangenesis
395
Contemporary precursors
Darwin was not the only nineteenth century writer to speculate on the
phenomena of heredity and reproduction. It may be that elements of his
hypothesis can be found in a number of contemporary works, but so far
only five nineteenth century men have been specifically credited with
espousing pangenetic hypotheses prior to Darwin.
Two of these men—Lorenz Oken and Paolo Mantegazza—did not at all
influence Darwin's developing conceptions. The evidence for Oken's commitment to a pangenetic hypothesis is a brief and unimpressive passage in
his Allegemeine Naturgeschichte (1839-1841; English edition, 1847). This
passage, quoted by Zirkle,97 is so tenuously connected with Darwin's highly
developed formulation of the hypothesis that Oken can be easily rejected as
a source of possible influence. Mantegazza's claim is based upon Darwin's
own unelaborated statement that Mantegazza in his Elementi di igiene 'clearly
foresaw' the hypothesis of Pangenesis.98 That is debatable, but unimportant,
since it is in any case highly unlikely that Darwin knew of this Italian work
until after he had seen Mantegazza's 1868 review of the Variation." There is
no evidence that Darwin ever read the appropriate passage in Oken's work,
and there is every reason to believe that he did not see Mantegazza's original
96. See Variation, n, 375, n. 29. Cf. Life and letters, n, 229. For a summary of the ideas of Buffon and
Bonnet, see Erik Nordenskiold, The history of biology, trans, from the Swedish by L. B. Byre (New
York, 1928), pp. 224-226, 244-246.
97. Zirkle (n. 20), p. 144.
98. See Darwin (n. 89), n, 359, n. 42.
99. Mantegazza's review appeared (in Italian) in Nuova Antologia, 1868, 8, 70-98. See esp. p. 97,
where Mantegazza claims a similarity between Darwin's ideas on Pangenesis and his own earlier
ideas, and where he documents this claim by citing the third edition of his Elementi di igiene exactly as
Darwin cites it in his Variation footnote. I suspect Darwin was attracted to read this review of his own
book, but had never read Mantegazza's earlier Elementi.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
pre-formed particles of Buffon and Bonnet and his own gemmules, which
'were not originally pre-formed, but are continually produced at all ages
during each generation, with some handed down from preceding generations.'96
Darwin could probably have demonstrated equally fundamental differences between his ideas and those of any of the pre-nineteenth century
pangenetic theorists. 'The provisional hypothesis of Pangenesis' was more
elaborate and more fully developed than heredity theories put forth before
1800 and Darwin undoubtedly developed his ideas independently of any
direct influence from these earlier theorists.
396
Journal of the History of Medicine : October 1969
statement before 1868, if ever. Even had he read both before drafting his
own 1865 manuscript, they would have given him little help.
However, three nineteenth century scientists—Owen, Spencer, and Naudin—probably did influence Darwin, each in different ways and to differing
degrees.
Richard Owen
100. Richard Owen, On parthenogenesis, or the successive production of procreating individuals from a
single ovum; a discourse introductory to the Hunterian lectures on generation and developmentfor the year 1849,
delivered at the Royal College of Surgeons of England (London, 1849).
101. More letters, 1, 102-103.
102. See T. H. Huxley, 'On the agamic reproduction and morphology of Aphis,' Trans. Linn. Soc.
Lond., 1863, 22, 193-220, 221-236. Cf. Nora Barlow, ed., Darwin and Henslow, the growth of an idea,
letters 1831-1860 (London, 1967), p. 209, n. 2.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
In his book On Parthenogenesis, published in 1849, Richard Owen stated
views in some respects similar to Pangenesis.100 Darwin probably read this
book soon after publication, and it is certain from a letter to Huxley in
1857 or 1858 that he was then familiar with Owen's hereditary hypothesis:
'I am extremely glad you have taken up the Aphis question, but, for Heaven's sake, do not come the mild Hindoo (whatever he may be) to Owen;
your father confessor trembles for you. I fancy Owen thinks much of this
doctrine of his; I never from the first believed it
' 101 The 'Aphis question,'
specifically the vegetative reproduction of Aphis, had received much attention in Owen's book; but Huxley's researches into the matter had convinced
him that both Owen's facts, and the hereditary theory by which he tried to
account for parthenogenesis, were seriously in error.102 This letter to Huxley, then, indicates not only that Darwin was familiar with Owen's 'doctrine' before drafting his Pangenesis manuscript of 1865, but also that he
agreed with Huxley's rejection of it, at least insofar as it applied to the
'Aphis question.'
Could a theory which he apparently rejected have exercised much influence on Darwin's developing conceptions? Obviously, it did not result
in Pangenesis full-grown. Darwin knew about Owen's theory before publishing the Origin; yet that book, as we have seen, contains nothing comparable to, or even reminiscent of, Pangenesis.
Nevertheless, Darwin may have been influenced by at least one aspect of
Owen's theory, namely Owen's concept of derivative germ cells, some of
which—instead of being used for the formation of the body—might 'remain unchanged and become included in that body':
Geison : Darwin and Pangenesis
397
. . . so included [wrote Owen], any derivative germ cell, or the nucleus of such,
may commence and repeat the same processes of growth by imbibition, and of
propagation by spontaneousfission,as those to which itself owed its origin.103
[Owen believes] that the primary germ-cell was formed within the ovarium of
the female and was fertilised by the male. My gemmules are supposed to be
formed, quite independently of sexual concourse, by each separate cell or unit
throughout the body, and to be merely aggregated within the reproductive organs.108
To Owen, already at odds with Darwin over natural selection, this difference did not seem fundamental: 'It may be a defect of power; but I fail,
after every endeavour, to appreciate the "fundamental difference" between
Mr. Darwin's cell-hypothesis of 1868 and mine of 1849.>109 In the second
edition of the Variation, Darwin faithfully reported this remark of Owen's,
but he was now able to enlist the support of a reviewer who had shown how
'really different' Pangenesis was from Owen's ideas.110 This reviewer—
besides pointing out that it was rather strange of Owen to claim priority for
a view which he now considered 'fundamentally erroneous'—suggested
that the only resemblance between Owen's theory and Darwin's was that
both assumed certain hypothetical gemmules; and he then showed how
irreconcilable Owen's gemmules, as the progeny of the primary impregnated germ-cell, were with Darwin's gemmules, which were 'thrown off
from the germinal matter of every cell throughout the body.' 111
103. Owen (n. ioo), p. 5; cf. p. 7.
104. Ibid., passim; see esp. p. 70.
105. See ibid., p. 25. Cf. Variation, n, 375, n. 29.
106. Variation, n, 383—385.
107. Ibid., n, 375, n. 29.
108. Ibid.
109. Owen (n. 85), in, 813.
n o . Darwin (n. 89), n, 359, n. 42.
i n . See J. Anat., Lond., 1869, j , 436-443, on pp. 441-442. The review is not of Darwin's Variation
but of Owen's then recent On the anatomy of vertebrates.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
Owen used these residual germ cells not only to explain his belief that
several generations of Aphis larvae could be produced from a single impregnated ovum 104 but also to account for parthenogenesis in general and for
regeneration and the repairs of injuries105—phenomena which Darwin was
later to explain on the basis of residual gemmules.106 Furthermore, as Darwin admitted in 1868, Owen's view agreed with his 'in the assumed transmission and multiplication of his germ cells. . . .' 107
But Darwin once again believed that there was a 'fundamental difference':
398
Journal of the History of Medicine : October 1969
Owen's ideas—more especially his concept of residual germ cells—may
well have influenced Darwin's gemmule hypothesis, but Owen's influence
alone could not have carried Darwin very far along the path toward his
highly developed formulation of Pangenesis.
Herbert Spencer
112. Herbert Spencer, Principles of biology (London, 1864-67), 2 vols. Hereafter cited as Principles.
113. Ibid., 1,177-183,25 3-256. It is from Spencer's preface that we know the date that each number
was issued to the subscribers. No. 9 (pp. 161-240) was issued in July 1863; No. 10 (pp. 241-320) in
January 1864. See ibid., 1, v.
114. Although no copy of the Principles of biology is listed in the catalogue of the Darwin library at
the Botany School, Cambridge (n. 92), Dr. H. Lewis McKinney has discovered in that same library
an annotated copy of the book; the separate subscriber's issues, together with their covers, are bound
up into one thick volume. Dr. McKinney has very generously allowed me to use his extensive notes
on Darwin's marginalia and annotations and has helped me secure from the Botany School at Cambridge a microfilm copy of the most important of these, so that I might see them with my own eyes.
Without Dr. McKinney's discovery and assistance, the discussion on Spencer would have been impossible.
115. On 30 June 1866 Darwin wrote to Hooker [More letters, n, 235): 'I have almost finished the
last number of H. Spencer, and am astonished at its prodigality of original thought. B u t . . . It is also
very unsatisfactory, the impossibility of conjecturing where direct action of external circumstances
begins and ends—as he candidly owns in discussing the production of woody tissues in the trunks of
trees on the one hand, and on the other in spines and the shells of nuts.' The discussion to which
Darwin refers appears on pp. 255-274 of Vol. n of Spencer's Principles, pages which were included in
the issue (No. 16) for June of 1866. So the letter to Hooker gives definite proof that Darwin read at
least one number of Spencer's Principles within a month after issue.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
Far more pangenetic in spirit than Owen's ideas were those which Herbert Spencer laid down in his Principles of Biology, sent periodically to
subscribers, and then published in two volumes in 1864 and 1867 respectively.112 Spencer introduced his pangenetic 'physiological units' in the
unbound issue of June 1863, and then, in the next issue—-January 1864—he
very specifically utilized them in a hereditary hypothesis.113 This latter issue
appeared more than a year before Darwin dispatched his manuscript of
Pangenesis to Huxley, so Darwin—who usually kept abreast of the contemporary literature—very probably read Spencer's ideas before formulating his own gemmule hypothesis. This is an extremely important point, and
one on which there is apparently no conclusive evidence. But since Darwin
was a subscriber to Spencer's Principles,114 and since there is evidence that he
was in the habit of perusing Spencer's numbers as they issued from the
press,115 it is almost certain that he read Spencer's hereditary hypothesis—
offthe press in January 1864—before drafting his own gemmule-hypothesis
in May 1865.
If Darwin did read Spencer before formulating his own ideas about Pangenesis, what he read there must have greatly influenced his own concep-
Geison : Darwin and Pangenesis
399
116. Principles, I, 177.
117. Ibid., 1, 177-182.
118. Ibid., I, 182-183.
119. See, e.g., Origin, p. 45, where Darwin describes individual differences as 'highly important'
because they 'afford materials for natural selection to accumulate.' Cf. Vorzimmer (n. 26), pp. 386390.
120. See Principles, 1, 256.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
tions. Spencer introduces his physiological units in a chapter on waste and
repair. He suggests that there are units in the circulating blood 'exactly like
in kind to those of which each organ consists,'116 and he makes these units
responsible for repair and regeneration by virtue of a 'faculty of organic
polarity.' This faculty, similar in many ways to Darwin's concept of'affinity,' was the abbreviated expression Spencer used for his assumption that
groups of compound units have the power to assimilate appropriate ambient materials and mould them into units of their own form. These organic
polarities are dependent upon the structure of the physiological units; and
by the total interaction of all polarities, the aggregate of physiological units
is compelled to take the form of the species to which they belong.117
Spencer calls these units 'physiological' because they are intermediate in
complexity between the 'chemical units' and the 'morphological units'
which make up organic bodies. Physiological units, that is to say, are far
more complex than the proximate organic compounds—albumin, fibrin,
even protein—but must be somewhat less complicated than cells, since
there are subcellular creatures (such as rhizopods) which nonetheless have
the power both to aggregate into a specific form and to perpetuate that
specific form in offspring.118
In his later chapter on heredity, Spencer expressed the belief that 'a positive explanation of Heredity is not to be expected in the present state of
Biology,' but he offered a hypothesis in which the physiological units are the
proximate agents of inheritance. For Spencer, the sex cells are but vehicles
for the physiological units, and 'the likeness of any organism to either
parent, is conveyed by the special tendencies of the physiological units
derived from that parent.' Each parent contributes a group of physiological
units, thus guaranteeing, in general, a blended inheritance. But, said Spencer, the transmission of'individual peculiarities,' or 'spontaneous' variations
is also thus guaranteed. (It was these 'individual differences' upon which
Darwin primarily relied for the raw material for selection.)119 Spencer also
suggested that the effects of use and disuse were transmitted by virtue of
modifications impressed upon the structures (and therefore upon the organic polarities) of the physiological units.120
400
Journal of the History of Medicine : October 1969
The similarities between these ideas and Darwin's later theory is striking.
In 1868 Darwin pointed out, in a long footnote in the Variation, that
Spencer's physiological units performed nearly all the functions and possessed nearly all the attributes of his own gemmules:
Darwin then cited Spencer's chapters on waste and repair and on heredity,
but he might also have cited Spencer's later chapter on variation, in which
variation is very specifically attributed to quantitative and qualitative differences in the physiological units which are aggregated into the sexual
elements of each parent, and in which Spencer speaks of the segregation of
these units as one explanation for the fact that offspring are never a perfectly
homogeneous mean between the two parents.122 Also germane is the chapter entitled 'Genesis, Heredity, and Variation,' in which Spencer recapitulates and extends some of his earlier remarks on the hypothesis of physiological units, and in which he concludes that 'the power of this hypothesis to
explain so many phenomena, and to bring under a common bond phenomena that seem so little allied, is strong evidence of its truth.' 123 Both of these
chapters appeared in the issue ofJanuary 1864.
But despite the remarkable similarities between Spencer's physiological
units and his own gemmules, Darwin thought he detected differences in
that Spencer had not applied his units specifically to reversion and in that a
certain number or mass of gemmules were required for the development of
each cell or part. 'Nevertheless,' continued Darwin in the 1868 edition of
the Variation, 'I should have concluded that Mr. Spencer's views were
fundamentally the same with mine, had it not been for several passages
which, as far as I understand them, indicate something quite different.'124
He then quoted from Spencer's Principles three passages which seemed to
him to indicate that the physiological unit hypothesis was not quite equivalent to the gemmule hypothesis. Two of these quoted passages involve
rather trivial differences, but the other passage—the longest—seems to indicate that Darwin and Spencer did disagree on one fundamental point. In
this passage, Spencer had written:
121. Variation, n, 375, n. 29.
122. Principles, 1, 266-268.
123. Ibid., 1, 291.
124. Variation, n, 375, n. 29.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
[Spencer's physiological units] agree with my gemmules in being supposed to
multiply and to be transmitted from parent to child; the sexual elements are supposed to serve merely as their vehicles; they are the efficient agents in all the forms
of reproduction and in the repairs of injuries; they account for inheritance . . . ;
they are supposed to possess polarity, or, as I call it, affinity, and apparently they
are believed to be derived from each separate part of the whole body.121
Geison : Darwin and Pangenesis
401
This certainly seems to be a direct contradiction of Darwin's assumption
that cells modified by use or disuse would give offcorrespondingly modified
gemmules. But Darwin failed to quote Spencer's subsequent statement
that, in the end, it was necessary to admit that 'ifthe structure of [any given]
organism is modified by modified function, it will impress some corresponding modification on the structures and polarities of its units.'126
After reading the chapter on Pangenesis in the Variation, Alfred Russel
Wallace wrote Darwin to tell him how much he admired it. 'You have now
fairly beaten Spencer on his own ground,' he wrote, 'for he really offered no
solution of the difficulties of the problem.' In a later letter, Wallace added
that though he was a great admirer of Spencer, he felt 'how completely his
view failed to go to the root of the matter, as yours does. His [view]
explained nothing. . . . Yours, as far as I can see, explains everything in
growth and reproduction....' Unfortunately, Wallace's critique of Spencer's hypothesis was never very specific; the closest he came to a detailed
analysis was in saying:
As I understood Spencer, his physiological units were identical throughout each
species, but slightly different in each different species; but no attempt was made to
show how the identical form of the parent or ancestors came to be built up of such
units.127
Spencer himself, after reading the chapter on Pangenesis, apparently told
Darwin that those views of his which Darwin had quoted in the footnote
'refer to something quite distinct,' and assured Darwin that the gemmule
hypothesis was 'quite different' from his own. 128 Spencer's concession inspired a most interesting remark by Darwin. In a letter to Hooker, he
wrote:'... this [Spencer's concession] is a great relief to me, as I feared to be
125. Principles, I, 255.
126. Ibid., l, 256.
127. More letters, I, 300-301.
128. Life and letters, n, 262, 260.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
It is not obvious that change in the form of a part, caused by changed action, involves such change in the physiological units throughout the organism, that these,
when groups of them are thrown off in the shape of reproductive centres, will
unfold into organisms that have this part similarly changed in form. Indeed, when
treating of Adaptation..., we saw that an organ modified by increase or decrease
of function, can but slowly so re-act on the system at large, as to bring about those
correlative changes required to produce a new equilibrium; and yet only when
such new equilibrium has been established, can we expect it to be fully expressed
in the modified physiological units of which the organism is built—only then can
we count on a complete transfer of the modification to descendants.125
402
Journal of the History of Medicine : October 1969
After reading any of his books, I generally feel enthusiastic admiration for his
transcendent talents, and have often wondered whether in the distant future he
would rank with such great men as Descartes, Leibniz, etc., about whom, however, I know very little. Nevertheless I am not conscious of having profited in my
own work by Spencer's writings. His deductive manner of treating every subject
is wholly opposed to my frame of mind. His conclusions never convince m e . . . .
His fundamental generalisations... are of such a nature that they do not seem to
me to be of any strictly scientific use. They partake more of the nature of definitions than of laws of nature. They do not aid one in predicting what will happen
in any particular case. Anyhow they have not been of any use to me.133
129. Ibid., n, 260-261.
130. Darwin (n. 89), n, 359, n. 42.
131. See Life and letters, n, 260-261, where Darwin reports that Spencer 'is not sure he understands
it. . . .'
132. Principles, 1, 255.
133. Barlow (n. 6o), pp. 108-109. My italics.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
accused of plagiarism, but utterly failed to be sure what he meant, so
thought it safest to give my view as almost the same as his.. . .' 129
More than anything else, it was probably these assurances from Spencer
himself which allowed Darwin to say with such confidence in the second
edition of the Variation, 'I formerly thought that the physiological units of
Herbert Spencer . . . were the same as my gemmules, but I now know that
this is not the case.'130 It is certainly curious that Spencer failed to recognize
the fundamental similarities between his views and those of Darwin, but
perhaps this failure can be traced to the same unknown difficulties which
apparently prevented him from fully understanding Darwin's Pangenesis131 (and which, on the other hand, prevented Darwin from being totally
certain what Spencer meant).
These difficulties may have resided in the very different intellectual styles
of the two. For Spencer was a remarkably original and primarily deductive
thinker, while Darwin was much more inclined to offer extensive observational confirmation of his speculations. The point can be illustrated by
reference to the question of the inheritance of the effects of use and disuse.
Spencer and Darwin ultimately arrived at the same conclusion—namely,
that to some extent at least, the effects of use and disuse are transmitted to
offspring—but they got there by different routes. While Darwin listed a
catalogue of recorded cases which pointed toward the conclusion, Spencer
argued that the conclusion appeared 'to be a deduction from first principles. . . ,' 132
Darwin was clearly aware of this difference between Spencer's approach
and his own. In his autobiography, he wrote of Spencer:
Geison : Darwin and Pangenesis
403
134. Darwin's copy of Spencer's Principles, i, 179-183 (see n. 114 above).
135. Ibid., 1, 254-255.
136. Ibid., 1, 279.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
The key words in this appraisal may be 'am not conscious.' Spencer's deductivism undoubtedly struck Darwin as inadequate and sometimes silly, but
there may well have been aspects ofhis thought—especially ofhis hereditary
hypothesis and of some ofhis ideas on the inheritance of acquired characters
—which Darwin incorporated almost unconsciously into his own thought
processes. Darwin's hereditary hypothesis was presented in a different manner and from a different point ofview; but the fact remains that most, if not
all, of the attributes of Darwin's gemmules belonged in the first place to
Spencer's physiological units.
From Darwin's subscriber's copy of Spencer's Principles, it is clear that he
was deeply impressed by many of Spencer's ideas on heredity and variation.
In marginalia and in notes affixed to the back page of his copy, Darwin
indicated those passages which struck him with special force. It is impossible
to know exactly when Darwin made these indications, but if (as seems
likely) he made at least most of them during hisfirstreading, then there can
be no doubt that his conception of Pangenesis owed much to Herbert
Spencer.
In the chapter on waste and repair, Darwin marked the passages in which
Spencer speaks of the regeneration of lost legs and tails in lizards and of lost
claws in lobsters, as well as the passage in which he introduces his physiological units. Beside a passage describing Spencer's concept of 'organic polarity'
Darwin wrote, 'like my attraction or selective [?] affinity.'134
Darwin made no marks in Spencer's next three chapters (entitled 'Adaptation,' 'Individuality,'and 'Genesis'),but in the chapter entitled 'Heredity,'
his interest is again revealed. The passages in which Spencer presents his
hereditary hypothesis are heavily marked. Beside these passages Darwin
wrote three marginal annotations, two of which say 'Pan,' undoubtedly
short for 'Pangenesis,' and the third of which reads: 'Think [?] he has not
got idea.'135
In Spencer's recapitulative chapter, 'Genesis, Heredity, and Variation,'
Darwin's marginalia suggest that his attention was drawn especially to
Spencer's concept that fertilization is the primary initiator of variation because it involves a mixture of the 'physiological units separated from an
adult organism... with the slightly-different physiological units of another
organism.'136 Darwin also marked the passage in which Spencer writes that
'the power of this hypothesis [of physiological units] to explain so many
404
Journal of the History of Medicine : October 1969
de novo.
However, it is also possible that Darwin made these annotations (though
perhaps not the marginal marks) after drafting the 1865 version of Pangenesis, at which time he might have returned to Spencer's work in search
of similarities and differences between his gemmule hypothesis and Spencer's physiological units. This interpretation seems the more likely because
in the Variation of 1868, Darwin quoted both of the passages marked 'Pan'
precisely to suggest that there were differences between his ideas and Spencer's.138 If this interpretation is accurate, Darwin may have incorporated
Spencer's ideas into his hypothesis of Pangenesis without full consciousness
of the debt he owed him. Perhaps only after re-reading Spencer did he
realize how similar their ideas were.
Charles Naudin
In i860 the French Academie des Sciences announced that the prix des
sciences physiques for 1862 would be awarded to the best essay submitted on
the topic of hybridization in the vegetable kingdom. Two essays were submitted, and the prize was given to Charles Naudin (1815-99), w n o had
begun his studies of plant hybridization in the 1840s with taxonomic problems in mind but had moved on in the next decade to consider their evolutionary significance.139 He was especially impressed by the tendency of
hybrid offspring to revert to the characters of one or the other of the pure
137. Ibid., 1, 291.
138. See Variations, n, 375, n. 29.
139. See Olby (n. 18), p. 62.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
phenomena, and to bring under a common bond phenomena that seem so
little allied, is strong evidence of its truth.' 137
Darwin made other marginal marks throughout much of the rest of the
book, but they have no direct connection with Pangenesis. They are interesting primarily because they demonstrate Darwin's consistent interest in
Spencer's ideas on the heritable effects of use and disuse. Spencer emphasized the effects of the environment and of use and disuse far more than
Darwin had, at least in the first edition of the Origin.
It would be particularly interesting to know when Darwin made the
marginal annotations 'Pan,' 'Like my attraction or selective [?] affinity,'and
'Think [?] he has not got idea.' If he made them in 1863 or 1864, when he
almost certainly first read Spencer's Principles, he must already have been
working Pangenesis out in his mind, so that Spencer's passages merely
reminded him of his own developing ideas. The passages marked 'Pan,' for
example, could hardly have suggested to Darwin the word 'Pangenesis'
Geison : Darwin and Pangenesis
405
parental species, a tendency which in 1858, if not before, he elevated to the
rank of a law, the 'law of return.' 140
Naudin's commitment to this law led him to the position that hybrids
were 'useless machineries' (rouages inutiles) which, unlike species, had no
place in the plan of Nature;141 and, as R. C. Olby points out, Naudin had
suggested as early as 1856 that reversion was accomplished in hybrid offspring via the segregation of the specific 'essences' which had been bequeathed to the hybrid:
In his prize essay, Naudin extended and emphasized this hypothesis to
explain reversion. The paper was not published in full until 1865, but the
second part appeared in 1863. a n d a review in 1864.143 Naudin and Darwin
apparently corresponded between 1862 and 1882,144 and we know that
Darwin received reprints of Naudin's 1858 paper and of the second part of
his prize paper.145 It is clear that Darwin had read this second part, which
contained the hypothesis for reversion, by September 1864.146
Darwin was much impressed by Naudin's 'excellent memoir,'147 and
found his hereditary hypothesis ofspecial interest because it bore specifically
on the question of reversion. Whenever he discussed hybridism in the
Variation, Darwin made wide use of Naudin's paper; and in the chapter on
Pangenesis, he was especially influenced by Naudin's notion of the segregation of the 'elements' or 'essences' of the two species which were crossed. In
140. See Charles Naudin, 'Observations concernant quelques plantes hybrides qui ont ixi cultiv£es
au Museum,' Ann. Sci. nat. (Bot.), 1858, sir. 4, g, 257-278, esp. pp. 277-278.
141. Ibid., pp. 277-278.
142. Charles Naudin, 'Observations constatant le retour simultane' de la descendance d'une plante
hybride aux types paternels et maternels,' C.JJ. Acad. Sci., Paris, 1856, 42, 625-628, on p. 628. As
translated by Olby (n. 18), p. 63.
143. (a) Charles Naudin, 'Nouvelles recherches sur l'hybriditi dans les v£g£taux,' m£moire pr£sent£
a 1'Academic des Sciences en Decembre 1861 et couronn6 dans la seance du 29 Decembre 1862,
Nouv. Arch. Mus. Hist, nat., Paris, 1865, 1, 25-174; (b) Charles Naudin, 'Nouvelles recherches sur
l'hybridit£ dans les vegetaux," deujrieme partie, Ann. Sci. nat. (Bot.), 1863, sir. 4, 19, 180-203; (c)
'Vegetable hybridity,' Nat. Hist. Rev., 1864, pp. 50-57.
144. See Olby (n. 18), p. 66.
145. Vorzimmer (n. 26), p. 383, n. 62, identifies Naudin's paper of 1858 as #161 in the Darwin
Reprint Collection, Botany School Library, Cambridge University; the second half of Naudin's
prize-winning essay is identified as #300. The citations which Vorzimmer gives these two papers is
misleading; cf. those given in n. 140 and n. 143 (b) above.
146. See More Letters, n, 339-340.
147. See, e.g., Variation, n, 386, n. 42 and 401, n. 54.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
Does one not say that Nature is eager to dissolve hybrid forms which do not enter
into her plan, and that she does this by the imperfection of the pollen in a large
number of hybrids, but also when these hybrids are fertile, by the separation of
the two specific essences which art or chance has violently brought together?142
406
Journal of the History of Medicine : October 1969
There is [wrote Darwin] hardly any greater difficulty in believing that these
many gemmules may unite or cohere, each retaining its own power and qualities, into
a single true germ, than in the well-known union of two species into a hybrid, and
of the hybrid with another hybrid until several species are commingled in a single
individual. Most of those who have closely studied hybrids and mongrels, especially M. Naudin, believe that all the characters of both parent-species are commingled, often in very unequal degrees in the unified product, but are notfused together or changed in nature like two elements in a chemical union.151
It is interesting indeed that this view of Naudin's confirmed a belief
Darwin had already 011857 or 1858 suggested tentatively to Huxley:
Approaching the subject from the side which attracts me most, viz., inheritance,
I have lately been inclined to speculate, very crudely and indistinctly, that propagation by true fertilization will turn out to be a sort of mixture, and not true fusion, of two distinct individuals, or rather of innumerable individuals, as each
parent has its parents and ancestors. I can understand on no other view the way in
which crossed forms go back to so large an extent to ancestral forms. But all this,
of course, is infinitely crude. . . .152
This celebrated passage caused Gavin de Beer to say that Darwin once
'came within a hair's breadth of making Mendel's discovery. . . . For that
148.
149150.
151.
152.
Variation, II, 386.
Cf. n. 143 (c), p. 55.
Naudin (n. 143, b), pp. 191-192; trans, into English by Olby (n. 18), pp. 167-168.
Olby (n. 27), p. 259. My italics.
More letters, 1, 103.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
accounting for those cases of crossed forms which exhibited the uncombined characters of either parent in 'stripes or blotches,' Darwin followed
Naudin's suggestion that the 'essences' or 'elements' ofeach species possess a
special affinity for their own kind.148
Naudin believed that a hybrid was a 'living mosaic' in which distinct or
pure essences from each parent co-existed in dynamic tension. Despite
appearances, no real 'fusion' of elements took place, so the parental elements
retained always their distinctive characters. Naudin even went so far as to
suggest that it was in pollen and ovules, especially pollen, that the tendency
to disunion was most energetic.149 For Naudin, so-called 'blending' inheritance was an illusion caused by our inability to detect the distinct units
whose mechanical, rather than chemical, union resulted in an apparent
blending ofparental traits.150 In the Pangenesis manuscript of 1865, Darwin
apparently accepted without qualification this concept of Naudin:
Geison : Darwin and Pangenesis
407
I have read with interest Bentham's address on hybridism. I am glad that he is
cautious about Naudin's view, for I cannot think that it will hold. The tendency of
hybrids to revert to either parent is part of a wider law (which I am convinced I
can show experimentally), namely, that crossing races as well as species tends to
bring back characters which existed in progenitors hundreds and even thousands
of generations ago. Why this should be so, God knows. But Naudin's view throws
no light, that I can see, on this reversion of long-lost characters.156
It may seem that Darwin was exaggerating this point since Naudin specifically suggested that reversion often progressed by 'insensible gradations,'
so that it might sometimes require a 'long series of generations' to achieve a
153. De Beer (n. 10), p. 208. Cf. R. A. Fisher, Tlte genetical theory of natural selection (New York,
1958), pp. 1-2. Fisher remarks of this passage that it indicates that Darwin once felt 'the need for an
alternative to blending inheritance . . . , though probably he never worked out a distinct idea of a
particulate theory. . . . The idea apparently was never developed . . . . " I would disagree: though
Darwin never did work out a general particulate theory, he also never lost sight (at least certainly not
after Naudin's paper of 1863) of the need for non-fusing, non-contaminating factors whenever he
considered the remarkable phenomenon of reversion. The 'Mendelian' character of Darwin's explanation
for reversion was emphasized forty years ago by H. F. Roberts in his important discussion of Darwin's
work on plant hybridization. Indeed Roberts argues that Darwin's explanation for reversion 'was a
natural corollary to his doctrine of pangenesis.' See H. F. Roberts, Plant hybridization before Mendel
(Princeton, 1929), pp. 221-240, esp. 238-240, quote on p. 240.
154. Variation, 11, 400-401.
155. See Olby (n. 18), p. 66.
156. More letters, n, 339-340.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
brief moment Darwin caught a glimpse of the particulate nature of inheritance and of the non-fusion, non-contamination of its factors. . . .' 153
In 1868 Darwin diverged to some extent from this position by suggesting
that so-called 'hybridised' gemmules—apparently a true fusion of the pure
parental gemmules—were responsible for the generally blended appearance
of hybrids. But even then he preserved the notion of pure parental gemmules whose development in a later generation would result in reversion to
the original parent-species trait.154 In truth, the general drift of Darwin's
ideas on inheritance was toward the doctrine of blending inheritance, but
whenever he considered the remarkable phenomenon of reversion, he recognized the need for a non-blending, or particulate, component in any
adequate hereditary hypothesis. Naudin's hypothesis for reversion in plant
hybrids found in Darwin a mind at least partially prepared.
For Darwin there was one major inadequacy in Naudin's hypothesis. In
the margin of his copy of Naudin's prize essay, he wrote simply: 'This view
will not account for distant reversion.'155 He referred to this objection in a
letter to Hooker, dated 13 September 1864:
408
Journal of the History of Medicine : October ig6g
cross had been lost.159
Darwin solved this problem by arguing that pure parental essences, or
gemmules, were preserved in a dormant, stable condition throughout a
long series of generations, thousands even. Since these pure, dormant gemmules were preserved and transmitted in an unmodified form, Darwin
suggested that when in later generations two hybrids were crossed, 'the
combination of pure gemmules derived from the one hybrid with the pure
gemmules of the same parts derived from the other, would necessarily lead
to complete reversion of character; and it is, perhaps, not too bold a supposition that unmodified and undeteriorated gemmules of the same nature
would be especially apt to combine.'160
It is this aspect of Darwin's hereditary ideas which Olby fails to give its
due. He does admit that 'the sheer weight of empirical evidence prevented
Darwin from ruling out non-blending heredity altogether and caused him
to incorporate Naudin's hypothesis of segregation into his theory of heredity in order to account for the behaviour of hybrids.'161 But in his desire,
generally justified, to associate Darwin with 'soft' or blending heredity,
Olby minimizes the extent to which Darwin accepted Naudin's hypothesis
for reversion and thereby obscures the nature of Naudin's influence on
Pangenesis.
Olby argues, in particular, that Darwin was especially eager to accept
Naudin's work just insofar as it supported his contention that there was no
basic difference between sexual and asexual reproduction,162 meanwhile
157. Naudin (n. 143, b), pp. 194-195.
158. Ibid., p. 195.
159. Variation, n, 49.
160. Ibid., n, 400-401.
161. Olby (n. 18), p. 66.
162. Ibid., pp. 93-94.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
complete return.157 But Naudin believed that in the great majority of cases,
there was at least some indication of return to parental type in the second
generation; indeed, he believed that some hybrids would return completely
as quickly as that. Even in his most extravagant example—the cross between
Nicotiana persica and Langsdorjjii—Naudin was thinking in terms often or
so generations,'158 while Darwin was thinking in terms of hundreds or
thousands of generations. According to Naudin's hypothesis, a dynamic
tension was created by the continuous struggle for dominance between the
pure elements of each parent and by the special affinity of each distinct
parental element for its own kind. Since this dynamic tension produced a
rather rapid reversion (Nature being 'eager' to dissolve the useless hybrids),
it was not clear to Darwin how Naudin's view could account for the remarkable ability of reversion to take place even after all traces of the original
Geison : Darwin and Pangenesis
409
Conclusion
Aside from the particular question of Naudin's influence, Olby has made
an impressive case for the view that Darwin's assumption of a basic identity
between sexual and asexual processes was crucial to his formulation of
Pangenesis. There can be no dispute on this point. It is amply confirmed by
reading either the 1865 or the 1868 version of Pangenesis, and Olby's analysis
ofthe 1865 manuscript demonstrates it with striking clarity. Olby may even
be correct in arguing that this assumption was the starting point for Pangenesis.165 Darwin's presentation of the hypothesis—especially in 1865, but
also in 1868—certainly seems to corroborate this view.
Nonetheless, Olby offers no explanation as to why Darwin was unable
to provide an hereditary hypothesis until 1865, even though he had long
before presented evidence for his belief that the various modes of reproduction pass into each other by insensible degrees.166 That Olby omits this
163. Ibid., p. 62; cf. pp. 66-67.
164. Variation, n, 48-49.
165. Olby (n. 18), pp. 99-100.
166. As early as 1848, for example, during his work on the barnacles, Darwin reported his conviction that 'hermaphrodite species must pass into bisexual species by insensibly small stages... .' (More
letters, 1,65.) And in 1859 he wrote that 'it is the opinion of most physiologists that there is no essential
difference between a bud and an ovule in their earliest stages of formation . . . ' (Origin, p. 10), a point
he later emphasized in the hypothesis of Pangenesis.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
largely ignoring the more 'Mendelian' feature of his work, namely his
hypothesis of specific segregation.163 Yet Darwin did not refer to Naudin's
work when discussing the identity of sexual and asexual processes, but
rather when discussing hybridism, and more especially, the tendency of
hybrids to revert to specific parental forms. Naudin's chief influence was to
show (or remind) Darwin that reversion could best be explained by assuming the existence of distinct, non-blending hereditary factors. It is true that
Darwin failed to advocate (though he did discuss) *64 Naudin's very Mendelian-like suggestion that these non-blending factors were segregated chiefly
in the sexual elements, but it cannot be doubted that he came away from
Naudin deeply impressed with the need for the transmission of stable, noncontaminating (i.e., Mendelian) hereditary units in order to explain reversion and allied phenomena (the sudden appearance of vestigial organs and
the emergence of'latent' secondary sexual characteristics). This influence
(and this influence alone) is clearly and specifically acknowledged in the
Pangenesis manuscript of 1865. Insofar as Naudin influenced the development of Darwin's hypothesis, he did so primarily along 'Mendelian' lines.
410
Journal of the History of Medicine : October ig6g
167. Olby (n. 18), p. 95.
168. See, e.g., Olby (n. 27), p. 254; Variation, n, 288-289, 291-292.
169. Hans Stubbe, Kurze Geschichte der Genetik bis zur Wiederentdeckung da Verebunsregeln Gregor
Mendels (Jena, 1965), p. 144.
170. See, e.g., Variation, n, 47.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
explanation is understandable in view of the fact that he is unaware of
Spencer's important influence and distorts Naudin's. The final crystallization of Darwin's thoughts on Pangenesis in 1865 depended on his reading of
Spencer's physiological unit hypothesis and on his finding in Naudin's
work an hypothesis to explain the phenomenon of reversion.
Secondly, Olby presents only one side of the story when he argues that
Darwin needed and wanted to assume that sexual and asexual reproduction
were essentially the same in order to explain 'how acquired characters can
be inherited by the sex cells and therefore how changes in the conditions of
life can giveriseto heritable variation in sexually reproducing organisms.'167
This argument is the subtlest of those in support of the view that Pangenesis
is a Lamarckian device to explain the inheritance ofacquired characters; but
it obscures the fact that Pangenesis was designed to explain not only heritable variation, but a variety of other phenomena, including reversion.
Darwin apparently considered the phenomenon of reversion as important
and 'wonderful' as the inheritance of acquired characters, and it was reversion which had brought him so close to 'hard' or non-blending heredity.
Darwin particularly emphasized the frequency of reversion in the asexual
process of budding ;168 and Hans S tubbe goes so far as to argue that Darwin's
writings on bud-variation demonstrate his reluctance, as late as 1868, to
rely on environmentally induced variation and his inclination to recognize
instead that 'the decisive cause of variability lay in the genetic constitution
of the organism.'169
Stubbe's 'defense' of Darwin's ideas on heredity seems even more onesided than Olby's attack on them. The truth lies somewhere between. In
1865, Darwin's was primarily a soft heredity, marked by blending and
environmental influences. But the phenomenon of reversion, and Naudin's
influence on Darwin's attempt to explain it, imparted to Pangenesis a
certain 'Mendelian' character. By 1868 this influence was largely muted
through Darwin's assumption that there existed 'hybridised' (i.e. fused or
blended) gemmules, and through his suggestion that the eventual development of even the pure (non-hybridized), dormant gemmules, depended
ultimately on the external conditions of life.170 But in 1865, when he first
articulated his hypothesis, Darwin apparently accepted without qualification Naudin's concept ofnon-blending hereditary units and also apparently
Geison : Darwin and Pangenesis
411
171. Olby (n. 27), p. 261.
172. See Vorzimmer (n. 26), pp. 386-390.
Downloaded from http://jhmas.oxfordjournals.org/ at University of Arizona on January 28, 2012
felt no need to make his dormant gemmules susceptible to environmental
conditions; in his manuscript of that year, he simply ascribed the eventual
development of the dormant gemmules to 'unknown causes.'171
Thus Darwin's first published explanation of reversion was, without
apparent qualification, a 'hard' one. Even though he allowed for the inheritance of acquired characters at the same time, Pangenesis was not simply a
'soft' hereditary hypothesis, and from the beginning was not designed to
account simply for heritable variation. If Darwin's later and fuller exposition of Pangenesis became more Lamarckian, this may have reflected his
gradually increasing reliance on a large number of individual variations172
and on Herbert Spencer's ideas about the heritable effects of use and disuse.
For the most part, the increasingly Lamarckian drift ofDarwin's thought on
heredity, rather than giving birth to Pangenesis, simply influenced his later
exposition of it.
Pangenesis, in short, was not a single-minded attempt to defend against
any attack Darwin's commitment to blending inheritance. Neither was it an
ad hoc attempt to account solely for the inheritance of acquired characters.
It was, rather, an attempt to develop a broad organizing principle, an attempt to connect under one point of view a wide range of previously
unconnected and poorly understood phenomena. As such, it had its origin
in what Geoffrey West called Darwin's 'naked need for order'—the same
need, it might be added, that gave birth to the lasting principle of natural
selection.