August 14, 1884] 



NA TURE 



369 



is produced, and so on indefinitely, and each hermaphro- 

 dite bud produces in succession an indefinite series of 

 similar buds. 



Now let us imagine a limit to this indefinite series of 

 buds, and examine its effect. 



r Suppose that, while B retains its power to produce in 

 succession an indefinite series of c's, the c's lose this 

 power. As the function of the ovary of C is to provide 

 the "generative blastema" for the ovaries of the series of 

 D's, and to mature the eggs E, which are to be fertilised 

 and developed within the bodies of the d's, it is plain that 

 with the loss by the c's of the power to reproduce by 



Em 



Em. 



Fig. 4 



budding, the ovary will be left without function, and we 

 should therefore expect it to disappear. C would then 

 become simply a male, but it would while young contain 

 a single unfertilised egg, EM, derived from the ovary of B. 

 If the power to bud were lost by the first generation of 

 buds, B, we should have the condition of things which is 

 shown in Fig. 2, where the hermaphrodite A produces a 

 male bud, B, and discharges the egg E into its body, 

 there to be fertilised and developed into the embryo EM. 

 A has, however, the power to repeat this process inde- 

 finitely, and to produce in succession a series of buds of 

 the generation B, and the life-history is therefore now 

 exactly shown in Fig. 3. 



F'g- 5 



Now suppose that, instead of appearing in succession, 

 a number of buds of the generation B are formed at the 

 same time, we shall then have the phenomena shown in 

 Fig. 4, where a number of eggs, E, E, e, e, are matured 

 simultaneously in the ovary, o, of the hermaphrodite A, and 

 are discharged while still unfertilised into the bodies of 

 the male buds, b, b, b, B, there to give rise to the embryos, 



EM, EM, EM, EM. 



This is very nearly what we have in Salpa, where very 

 many chain Salpa; are produced at one time. As these 

 have no power to reproduce by budding, they have lost 

 their ovaries, although each of them, when it is born, con- 



tains, like the buds of Pyrosoma, a single unfertilised egg, 

 derived, according to my observations and those of 

 Kowalevsky, and according to the analogy of Pyrosoma, 

 from the ovary of the solitary Salpa A, Fig. 5. 



The solitary Salpa is therefore a true female, and as it 

 has lost the testis which, according to the analogy of 

 Pyrosoma, its ancestors must have possessed, it is a true 

 female and not a hermaphrodite. 



We therefore have in place of the indefinite series of 

 hermaphrodite buds of Pyrosoma, a single generation of 

 male buds, each of which receives, like the buds of Pyro- 

 soma, a single mature ovum from the ovary of its gemmi- 

 parous ancestor. 



While the series of stages which are here described 

 may not correspond exactly to the actual phylogeny of 

 Salpa, it certainly shows that our knowledge of Pyrosoma 

 cannot be quoted as an " indirect negation " of my view ; 

 for it shows that the analogy of Pyrosoma would lead us 

 to look to the ovary of the solitary Salpa for the origin of 

 the egg which is fertilised and developed within the body 

 of the chain Salpa. 



All observers agree that the so-called " ovary " of the 

 chain Salpa consists of a single egg, which is fertilised 

 while the animal which carries it is very young and al- 

 most embryonic, and all agree that normally no more 

 eggs are produced by the chain Salpa. A " rudiment of 

 an ovary, which only consists of one fully developed 

 ovum " is certainly not an ovary at all, but an ovum, for 

 the origin of which we must search elsewhere, and 

 Kowalevsky's observation as well as my own show that it 

 originates in the ovary of the solitary Salpa. 



As all observers agree that the chain Salpa has a testis, 

 and that it normally contains only one egg, it is certainly 

 a male ; and as all observers agree that the solitary Salpa 

 has no testis, while Kowalevsky and I agree that it has 

 an ovary, we must regard it as a female, and we therefore 

 have, instead of an instance of alternation, a very remark- 

 able example of sexual difference. 



There seems to be only one way to escape this conclu- 

 sion ; — that is, by denying that the structure which 

 Kowalevsky and I have described as an ovary is an ovary 

 at all. 



Salensky is an advocate of this view, and claims that 

 my so-called ovary is simply a mass of embryonic cells 

 destined to give rise to the branchial sacs, as well as the 

 ovaries of the chain Salpae, but I have shown in a recent 

 paper in the Zoologisclier Anzeiger that the branchial 

 sacs of the chain Salpa originate from quite a different 

 part of the stolon, and that the ovary contains cells which 

 are in no sense embryonic or unspecialised, since they 

 have all the characteristics of ova. In a paper which is 

 now ready for publication, I shall give photographs of 

 sections which prove this point beyond question. 



R. N. G. gives, as another reason for rejecting my view, 

 the fact that Salensky has found a second ovary in a 

 chain Salpa. 



This is clearly exceptional, for all observers agree that 

 no such second ovary normally occurs, nor has Salensky 

 given conclusive proof that the cells which he observed 

 were ova at all, as he has not observed their development. 

 Out of many thousand sections which I have examined, 

 I have found three chain Salpa; which had received two 

 ova from the ovary of the solitary Salpa instead of one, 

 and if Salpa is descended from a form like Pyrosoma, it 

 is quite possible that a chain Salpa may occasionally re- 

 ceive with its ovum part of the ovary, and that this may 

 give rise to other ova, but the discovery of such an abnormal 

 Salpa would not prove that the normal chain Salpa is 

 hermaphrodite, even if it could be shown that these eggs 

 completed their development and became embryos. 



In conclusion, I wish to point out to R. N. G. that, 

 inasmuch as the writer who attempts to generalise from 

 the observed phenomena of science for the benefit of the 

 public should use every precaution to insure accuracy in 



