So 



NATURE 



[November 24, 1892 



larvae ( Veligers with large yolk-sac) three years ago, which first 

 afforded me a key to the solution of the problem of the relation 

 of the larva to food-yolk. 



The Arthropoda are an important group, for larval forms 

 widely prevail, especially, as is generally admitted, in the lower 

 forms. 



The nature of the Nauplius is too big a question for discussion 

 here, but Dohrn's conclusion that it is a transformed worm 

 larva {i.e. one with Arthropod characters) appears to me to 

 represent the truth. Two of the laws governing developmental 

 processes appear to be that larval organs may be transferred to 

 the service of the adult, and (more usually) that adult organs 

 may become larval, or, as thay may be termed, adaptational 

 larval organs. Numerous instances of both these could be 

 cited, and the three pairs of appendages in the Nauplius furnish 

 us with a case in point. 



In the development oi Mysis there is an example, of the con- 

 version of the Nauplius larva, typically represented in the 

 allied Euphansia, into a blastoderm upon a yolk-sac. 



I venture to attach most weight to the application of the prin- 

 ciple to the Vertebrala, for it is there that my own work has chiefly 

 lain, and it is undoubtedly the obstacles offered by the phenomena 

 of Vertebrate development which have hitherto prevented the 

 enunciation of the " law of development as an alternation of 

 generations." Larvae are so commonly encountered among the 

 Invertebrata that the wonder is that no one has inquired why 

 they are so rare, in any guise, in the Vertebrata. 



In this latter division of the animal kingdom it becomes 

 necessary to approach the problem previously stated as to the 

 fate of the larva when uterine development is initiated. 



It may firstly be noted that larval forms, equipped with 

 many adaptational larval organs, are to be encountered in cases 

 with complete segmentation and but little food-yolk, e.g. Mar- 

 sipobranchii, Ganoidei, and most Amphibia, while a blastoderm 

 on a yolk-sac is characteristic of Elasmobranchii, Teleosiii, and 

 Sauropsida, in which a larva, according to the common accepta- 

 tion, would not be very obvious. 



In all these cases, however, larval organs can be proved to 

 exist, and, most important of all, there is a well-marked larval 

 nervous system, which, while not certainly known to persist in 

 any adult form, has been proved to degenerate during the 

 ontogeny of all the os'v^2iXo\x's, Ichthyopsida as yet studied. This 

 apparatus is certainly neither a part of the adult nervous system 

 nor homologous with the latter. For an account of this 

 mechanism the reader may be referred to the papers cited below. ^ 



Among other larval structures referred to when reading the 

 paper, the curious degenerating cells on the blastoderm of 

 Pristiuriis, to which Prof. Van VVijhe once drew my attention, 

 and the knob on the blastoderm of Torpedo, as shown in Ziegler's 

 beautiful models of the embryos of this form, and as described 

 by H. E. and F. Ziegler in the Archiv fiir Mikroscopischc 

 Anatomie, Bd. xxxix. p. 85, deserve mention. 



The yolk-sac viewed as part of the larva would require detailed 

 and extended consideration. 



It is gradually broken down by some ferment action on the 

 part of the so-called merocytes,'-' which may possibly represent 

 degenerating cells of the larva. Only towards the close of life 

 in the egg-capsule does the yolk appear to be digested by the 

 alimentary tract of the Elasmobranch. In some reptiles, accord- 

 ing to Hans Virchow, the remains of the yolk-sac would appear 

 to be cast off. 



In the discussion on my paper, one speaker, a personal friend, 

 hinted that I had been led to look upon the yolk-sac as part of 

 the larva from having followed some stray ends of " larval nerve 

 fibres " on to that structure. I had to confess my regrets that at 

 that time I was unable to lay claim to any such observation ; in- 

 deed, that having cut my embryos ol Raja batis without any part 

 of the yolk-sac appended, it had never occurred to me that the 

 fibres described might pass to the yolk-sac. Quite recently it 

 has been seen that at any rate some of the larval " subepiblastic 

 nerves "of the Anat. Anz. paper do undoubtedly make their 

 zvay to the surface of the yolk-sac, lying just beneath its epiblastic 

 covering. That a further confirmation of my conclusions is to 

 be found in this observation goes without saying. 



I J. Beard: " The Transient Ganglion Cells and their Nerves ia .ffa/a- 

 batis. Anat. Anz., 1892, pp. 191-206; and also, " The Early Development 

 oi Lepidosteiis osseus," Proc. Roy. Soc. London, vol. xlvi. 1889, pp. 115- 



2_ It is to H. E. Ziegler that we owe most of our knowledge of the way in 

 which these merocytes, in their own degeneration and death, cause the ele- 

 ments of the yolk to become fit for absorption and assimilation. 



NO. 1204, VOL. 47] 



Some three years ago, when considering the " Inter- relation- 

 ships of the Ichthyopsida," at a time when this larval question 

 was prominently before me as a fascinating puzzle, I thought 

 that the larva disappeared above the Ichthyopsida. I was led 

 to this conclusion by reliance on the accepted belief that larvae 

 are only met with in aquatic animals, more particularly in marine 

 forms, and by the apparent absence of a larval nervous system 

 above the Ichthyopsida. My recent studies and the work of 

 Froriep and Robinson have taught me that this was erroneous. 



The larva never disappears, however much it may undergo 

 degeneration. 



It may even be doubted if there are not traces of the nervous 

 system of the larva in the ontogeny of the Amniota, for there 

 appear to be certain observations of Froriep on reptiles which 

 may admit of interpretation in this sense, and my friend. Dr. 

 Arthur Robinson, tells me that he believes he has found traces 

 of it in certain Mammalian embryos. In mammals, as will be 

 seen, the larva must be regarded as an internal parasite, and like 

 such it would yield up its chief organs. Some remains of its 

 nervous system may, however, persist, as I have proved to be 

 the case in Mustelus vulgaris, where the larva is almost as 

 parasitic as in the Mammal. The Amnion of the higher Ver- 

 tebrata is probably also a larval structure with analogies to the 

 organ of the same name in Insecta, in the Pilidium development, 

 &c., as Kennel had previously insisted. It would appear to me 

 to be a membrane conditioned by the way in which the adult is 

 formed upon the larva. 



Another important larval structure is the yolk-sac placenta of 

 Mustelus IcBvis and of many mammals. 



In the latter the importance of this organ during a long 

 period of foetal life has been proved by Hubrechtand Robinson. 



The yolk-sac placenta may be explained as due to the fixation 

 of a parasitic larva ; indeed, in mammals the larva has become a 

 fixed internal parasite in the uterus, and its mode of life, like 

 that of all internal parasites, leads to great degeneration. 



In this connection it may be insisted that it would be con- 

 trary to all that we know concerning the effects of the parasitic 

 mode of life to suppose that a form might become a fixed 

 internal parasite, and subsequently becoming freed from its host, 

 attain to a higher grade of organization. Yet this is what we 

 must believe to hold good, if the current views of mammalian 

 development be accepted as correct. From my standpoint, on 

 the contrary, the larva may become a fixed internal parasite, and 

 none the less there may arise upon it a more highly organized 

 and, when fully developed, free-living form, the Mammal. 



Witness must be borne to the circumstance that Miiller, 

 Kleinenberg, and Kennel have already recognized that in some 

 few divisions of the Invertebrata the mature form always arises 

 upon a larva. 



In such groups as the Echinoderms an alternation of genera- 

 tions is now an obvious explanation of the facts, and when so 

 magnificent an investigator as Johannes Miiller proved this 

 nearly fifty years ago, one asks, in vain perhaps, why modern 

 embryologists, like Korschelt and Heider in their otherwise 

 admirable " Entwickelungsgeschichte," ignore it. The 

 "recapitulation theory," and the question concerning the 

 nature of the mesoderm have overshadowed the fact and con- 

 cealed the recognition of an akernation of generations. But the 

 so-called "law of ontogeny" itself is no explanation of the 

 riddles of embryology ; at most the recapitulation hypothesis 

 holds for the development of organs, not of organisms. 



So far as the facts are available, Metazoan development 

 appears to me to be by means of an alternation of generations, 

 in that from the fertilized egg there arises an organism, the 

 larva, upon which, in one way or another according to the cir- 

 cumstances of each case, a new form, the adult or imago, takes 

 its origin. 



In 1855 the veteran zoologist, P. J. Van Beneden, wrote : — 

 ' ' La generation alternante est un phenomenc cju'ilfant chercher 

 a faire rentrer dans la loi commune de la reproduction et non 

 pas laisser comme une exceptio7i dans la science." 



In this essay an attempt has been made for the first time to 

 prove that it is "la loi commune de la reproduction" in 

 Metazoa,' and in concluding I cannot do better than echo the 

 beautiful aphorism of Goethe, which in a similar connection has 

 already been commented upon by Steenstrup and Von Baer : — 



''Die Natur geht ihren Gang, und was tins als Atisnahme 

 erschtint ist in der Kegel." J. Beakd. 



I It is not assumed that all the phenomena classified as " alternations of 

 generations " are alike in their nature. 



