September ii, 1890] 



NATURE 



473 



the nature of their food and the mode of obtaining it ; and the 

 transition from one stage to the other is not a gradual but 

 an abrupt one, at any rate so far as external characters are 

 concerned. 



Sudden changes of this kind, as from the free-swimming 

 Pluteus to the creeping Echinus, or from the sluggish leaf-eating 

 caterpillar to the dainty butterfly, cannot possiily be recapitu- 

 latory, for even if small jumps are permissible in Nature, there 

 is no room for bounds forward of this magnitude. Cases of 

 abrupt metamorphosis may always be viewed as due to secondary 

 modifications, and rarely, if ever, have any significance beyond 

 the particular group of animals concerned. For example, a 

 Pluteus larva may be recognized as belonging to the group of 

 F.chinoidea before the adult urchin has commenced to be formed 

 within it, and the Lepidopteran caterpillar is already an unmis- 

 takable insect. Hence, for the explanation of the meta- 

 morphoses in these cases it is useless to look outside the groups 

 of Echinoidea and Insecta respectively. 



Abrupt metamorphosis is always associated with great change 

 in external form and appearance, and in mode of life, and very 

 usually in mode of nutrition. A gradual transition in such cases 

 is inadmissible, because in the intermediate stages the animal 

 would be adapted to neither the larval nor the adult condition ; 

 a gradual conversion of the biting mouth parts of the caterpillar 

 to the sucking proboscis of a moth would inevitably lead to 

 starvation. The difficulty is evaded by retaining the external 

 form and habits of one particular stage for an unduly long period, 

 so that the relations of the animal to the surrounding environment 

 remain unchanged, while internally preparations for the later 

 stages are in progress. Cinderella and the princess are equally 

 possible entities, each being well adapted to her environment. 

 The exigencies of the situation do not permit, however, of a 

 gradual change from one to the other : the transformation, at 

 least as regards external appearance, must be abrupt. 



Kleinenberg has recently directed attention to cases in which 

 the larval and adult organs develop independently ; the larval 

 nervous system, for instance, aborting completely and forming 

 no part of that of the adult. I am not sure that I fully under- 

 stand Kleinenberg's argument, but it seems very possible that 

 such cases, which are probably far more numerous than is yet 

 admitted, may be due to what may be termed the telescoping of 

 ancestral stages one within another, which takes place in actual 

 development, and may accordingly be grouped under the head 

 of developmental convenience. Undue prolongation of an early 

 ancestral stage, as in cases of abrupt metamorphosis, must involve 

 modification, especially in the muscular and nervous systems ; 

 in such cases a telescoping of ancestral stages takes place, as 

 we have seen, the adult being developed within the larva. 

 Such telescoping must distort the recapitulatory history, and as 

 the shape of the larva and adult may differ widely, an indepen- 

 dent origin of organs, especially the muscular and nervous 

 systems, may be acquired secondarily. 



The stage in the development of Squilla, in which the three 

 posterior maxillipedes disappear completely, to reappear at a 

 later stage in a totally different form, is not to be interpreted as 

 meaning that the adult maxillipedes are entirely new structures 

 unconnected historically with those of the larva. Neither is the 

 annual shedding of the antlers of deer to be regarded as the 

 repetition of an ancestral hornless condition intercalated his- 

 torically between successive stages provided with antlers. ]n 

 both cases the explanation is afforded by convenience, whether 

 of the embryo or adult. 



Many embryological modifications or distortions may be attri- 

 buted to mechanical causes, and may fairly be considered under 

 the head of developmental conveniences. 



The amnion of higher vertebrates is a case in point, and is 

 probably rightly explained as due in the first instance to sinking 

 or depression of the embryo into the yolk, in order to avoid 

 distortion through pressure against a hard unyielding eggshell. 

 A similar device is employed, presumably for the same reason, 

 in the early development of many insect embryos ; and the 

 depression of the Taenia head within the cyst is a phenomenon 

 of very similar nature. 



Restriction of the space within which development occurs 

 often causes displacement or distortion of organs, whose growth, 

 restricted in its normal direction, takes place along the lines of 

 least resistance. The telescoping of the limbs and other organs 

 within the body of an insect larva is a simple case of such dis- 

 tortion ; and a more complicated example, closely comparable 

 in many ways to the invagination of the Taenia head, is afforded 



NO. 1089, VOL. 42] 



by the remarkable inversion of the germinal layers in rodents, 

 first described by Bischoff in the guinea-pig, and long believed 

 to be peculiar to that animal, hut subsequently and simulta- 

 neously discovered by three independent observers — Kupffer, 

 Selenka, and Fra«er — to occur in varying degrees in rats, mice, 

 and in other rodents. 



One of the most recent attempts to explain developmental 

 peculiarities as due to mechanical causes is Mr. Dendy's sugges- 

 tion with regard to the pseudogastrula stage in the development 

 of the calcareous sponges. It is well known that, while the 

 larva is in the amphiblastula stage, and still embedded in the 

 tissues of the parent, the granular cells become invaginated 

 within the ciliated cells, giving rise to the pseudogastrula stage. 

 At a slightly later stage, when the larva becomes free, the in- 

 vaginated granular cells become again everted, and the larva 

 spherical in shape ; while still later invagination occurs once 

 more, the ciliated cells being this time invaginated within the 

 granular cells. The significance of the pseudogastrula stage has 

 hitherto been undetermined, hut Mr. Dendy points out that the 

 larva always occupies a definite position with reference to the 

 parental tissues ; that the ciliated half of the larva is covered by 

 a soft and yielding wall, while the opposite half, composed 

 of the granular cells, is covered by a layer stiffened with rigid 

 spicules ; and his observations on the growth of the larva lead 

 him to think that the pseudogastrula stage is brought about me- 

 chanically by flattening of tiie granular cells through pressure 

 against this rigid wall of spicules. 



Embryology supplies us with many unsolved problems, and it 

 is not to be wondered at that this should be the case. Some of 

 these may fairly be spoken of as mere curiosities of develop- 

 ment, while others are clearly of greiter moment. I do not 

 propose to catalogue these, but will merely mention two or 

 three which I happen to have recently run my head against, and 

 remember vividly. 



The solid condition of the oesophagus, in Elasmobranch em- 

 bryos, first noticed by Balfour, is a very curious point. The 

 oesophagus has at first a well-developed lumen, like the rest of 

 the alimentary canal ; but at an early period, stage K of Bal- 

 four's nomenclature, the part of the oesophagus overlying the 

 heart, and immediately behind the branchial region, becomes 

 solid, and remains solid for a long time, the exact date of 

 reappearance of the lumen not being yet ascertained. 



Mr. Bles and myself have recently noticed that a similar 

 solidification of the oesophagus occurs in tadpoles of the common 

 frog. In young free-swimming tadpoles the oesophagus is per- 

 forate, but in tadpoles of about ']\ mm. length it becomes solid 

 and remains so until a length of about ic^ mm. has been at- 

 tained. The solidification occurs at a stage closely correspond- 

 ing with that in which it first appears in the dogfish, and a 

 curious point about it is that in the frog the oesophagus becomes 

 solid just before the mouth opening is formed, and remains solid 

 for some little time after this important event. 



This closing of the oesophagus clearly cannot be recapitula- 

 tion, but the fact that it occurs at corresponding periods in the 

 frog and dogfish suggests that it may possibly, as Balfour hinted, 

 "turn out to have some unsuspected morphological bearing." 



Another developmental curiosity is the duplication of the gill- 

 slits by gro Ath downwards of tongues from their dorsal margins ; 

 a duplication which is described as occurring in Amphioxus and 

 in Balanoglossus, but in no other animal ; and the occurrence of 

 which, in apparently closely similar fashion, is one of the 

 strongest arguments in favour of a real affinity between these 

 two forms. It is hardly possible that such a modification should 

 have been acquired independently twice over. 



A much more licigious question is the significance of theneur- 

 enteric canal of vertebrates, that curious tubular communication 

 between the central canal of the nervous system and the hinder 

 end of the alimentary canal that is conspicuously present in the 

 embryos of lower vertebrates, and retained in a more or less 

 disguised condition in the higher groups as well. 



Ihe neurenteric canal was discovered by that famous embryo- 

 logist Kowalevsky in Ascidians and in Amphioxus. He drew 

 special attention to the occurrence of a stage in both Ascidians 

 and in Amphioxus in which the larva is free-swimming and in 

 which the sole communication between the alimentary cavity 

 and the exterior is through the neurenteric canal and the central 

 canal of the nervous system ; and suggested ^ that animals may 



' •Weitere Studien aber die Entwickelungs-Geschichte dcs Am/kioxus 

 lanceolatus" {Archiv far mikrjskopische AnatomU, Bd. xiii., 1877, 

 p. 201). 



