120 Transactions of the Boyal Microscopical Society. 



the genninal vesicle of a mammalian ovum, it might contain above 

 five hundi'ed millions of gemmules. If these were lost or fully 

 developed at the rate of one in each second, this number would not 

 be exhausted until after a period of seventeen years. There would 

 thus be no difficulty in understanding why the characters of the 

 female parent might remain during life, even though apparently 

 dormant for many years. This is still more the case if we take 

 into consideration the entire ovum, since calculating on the sup- 

 position of its being a sphere y^^ of an inch in diameter it might 

 contain so many gemmules that if one were lost or developed in 

 each second they might not all be exhausted until after 5600 

 years. 



These calculations are made on the supposition that the entire 

 mass is composed of gemmules. Of this there is little probability ; 

 but still, even if a considerable portion of the ovum consists of 

 completely formed material and of mere nutritive matter, it may yet 

 contain a sufficient number of gemmules to explain all the facts 

 contemplated by the theory of pangenesis. The presence of any 

 considerable amount of such passive matter in the spermatozoa 

 would certainly be a serious difficulty in the way of the theory, 

 unless indeed a very considerable number are invariably concerned 

 in producing fertilization. 



"When, however, we come to apply similar reasoning to the 

 inheritance by the second or following generations of characters 

 which have remained apparently dormant in one or more previous 

 generations, it appears to me that the gemmule theory would fail, 

 unless gemmules have the power of reproducing others more or less 

 closely resembling themselves, and of collecting together more 

 especially in the sexual elements. This will, I think, be apparent 

 from the following considerations. 



An animal weighing 8 stones would contain about 3000 cubic 

 inches, and thus its entire volume would be about six millions of 

 millions times that of the germinal vesicle of an ovum. Hence, if 

 the number of gemmules in a vesicle as given above were present 

 in the grown-up animal and equally distributed over the whole 

 body, there would only be enough to allow one for each thousand 

 ova, or only one for a much greater number of spermatozoa. 



I have treated this question entirely in its physical aspect, and 

 made no reference to any other class of facts. The conclusions to 

 which I have been thus led agree remarkably well with those of 

 Darwin, though drawn from entirely difierent data. As will be 

 seen, the probable size of the ultimate molecules of Hving matter is 

 sufficiently minute to make the gemmule theory possible when 

 examined fi'om a purely physical point of view. If there had been 

 good evidence to prove that the ultimate atoms of matter are 

 very much larger than indicated by the properties of gases, the 



