February 24, 1905.] 



SCIENCE. 



289 



that there is a particular substance or 

 'idioplasm' peculiar to each species of 

 plant or animal that is transmitted in the 

 germ-cells and has the power to determine 

 the development of the egg according to its 

 nature. Later research has given very 

 strong reason to accept this view in prin- 

 ciple, and for the further conclusion that 

 this physical basis is represented by a 

 substance contained within the nucleus and 

 known to cytologists as ' chromatin. ' Pass- 

 ing over the cogent, and I believe steadily 

 accumulating, evidence on which this con- 

 clusion rests, let us ask how the idioplasm 

 is to be conceived. Some of those who have 

 accepted the general conception of the idio- 

 plasm have endeavored to think of it as a 

 very complex but still single and homo- 

 geneous substance— the frog's egg, for ex- 

 ample, might be conceived as containing 

 a frog-determining substance, the human 

 germ a man-determining substance, and so 

 on. The most recent researches are, how- 

 ever, continually strengthening the ground 

 for a quite different conception, indicating 

 that the chromatin does not operate as a 

 simple substance, but is built into a com- 

 plex fabric having a definite architecture. 

 We are not here concerned with the par- 

 ticular form of this conception developed 

 by Weismann in his well-known work on 

 the Germ-plasm, and elsewhere. I am re- 

 ferring to more recent results of observa- 

 tion and experiment which are giving new 

 and more concrete evidence that the nu- 

 cleus possesses a complex organization, and 

 apparently one that must be conceived as a 

 kind of primary or original preformation, 

 which bears a certain analogy to that as- 

 sumed by Bonnet, though quite distinct 

 from it. 



"We may perhaps most readily approach 

 the grounds for this conclusion by consid- 

 ering, first, an example of the indirect evi- 

 dence drawn from recent experiments on 

 inheritance. I give a single example. 



typical of a large number of known cases, 

 of the heredity of single or unit characters 

 in the so-called Mendelian inheritance. If 

 pure gray mice be crossed with pure white 

 albino forms, the hybrid of?spring are all 

 gray without visible trace of white. But 

 if these gray hybrids be now paired with 

 each other, both parents being gray, ap- 

 proximately 25 per cent, of their progeny 

 are pure white without a trace of gray, and 

 they continue to produce pure white off- 

 spring thereafter. Many similar cases are 

 known, . the same proportion of approxi- 

 mately 25 per cent, of the 'recessive' char- 

 acter in the third generation holding true, 

 sometimes with great precision. What 

 does this prove? First, that the white 

 character is not really absent in the gray 

 hybrids hut only masked or concealed — 're- 

 cessive, ' in Mendel 's terminology ; secondly, 

 that the latent white character may in 

 the following generation be completely dis- 

 entangled or extracted from the gray; 

 thirdly, since the proportion is definite, that 

 the extraction takes place by means of some 

 definite mechanism. We are at present, 

 I think, unable to imagine an explanation 

 of these truly astonishing facts save by the 

 assumption that the gray and w^hite char- 

 acters are borne in the egg by correspond- 

 ing discrete bodies or entities of some kind, 

 that may be mixed and unmixed without 

 fusion, shuffled and unshufifled like cards in 

 a pack. The evidence is so far wholly in- 

 direct, though I think none the less cogent. 

 But now, bearing in mind that the ease of 

 the gray and white mice is but a single 

 example of a widespread phenomenon, let 

 us ask whether we can actually find any 

 definite structures in the egg, and particu- 

 larly in the nucleus, that may be assumed 

 to represent such entities. One of the most 

 significant and remarkable discoveries of 

 modern biology is the fact that such enti- 

 ties exist, though it is important not to 

 forget that their significance in heredity is 



