26o 



NATURE 



[Jan. lo, 1889 



sense-organs of these Annelids are the only ones of which 

 we know they in any way fulfil the conditions required of the 

 ancestry of the lateral sense-organs of Vertebrates, It appears 

 also that, if we admit the proposed homology as based upon 

 the Capitellidse, we can carry the origin of the lateral sense- 

 organs of both groups back to still simpler structures, for it 

 seems clear that the lateral sense-organs of Annelids have 

 been derived from cirri and portions of the parapodial ganglia 

 (Kleinenberg, Eisig). The rest of the Vertebrate sense-organs 

 are easily accounted for. It is becoming more and more prob- 

 able that the nose and ear are modified portions of the system 

 of lateral sense-organs, and I am not without hopes of showing 

 that the taste-buds of the higher Vertebrates may be derived 

 from lateral sense-organs which wandered through certain of the 

 gill-clefts during the development. 



The eyes are not difficult to account for, as plenty of Annelids 

 have eyes at the extreme anterior end of the ventral cord, in 

 connection with the first ganglion. 



To refer, in passing, to another structure. The Vertebrate 

 notochord has been shown by Ehlers and Eisig to correspond 

 with the ^' Nebendarm" of Annelida. At the risk of being 

 pronounced a heretic, I venture the opinion that the swimming- 

 bladder of fishes is also a derivative of the " Nebendarm," and 

 that the notochord and swimming-bladder are parts of the 

 same structure which have acquired different functions, and so 

 developed divergently. It is certainly not difficult to raise 

 arguments against the so universally accepted homology of 

 swimming-bladder and lungs. 



And now let me refer to some recent results of my own on the 

 Vertebrate nervous system. I have depicted them, very dia- 

 grammatically, in the accompanying figure. It represents a 



■N. Ac 



Diagrammatic transverse secticn through the trunk region of a Vertebrate 

 embryo. ii>, epiblast ; Z, Zwischenstrang oi Prof. His; Sp.Gl. 

 spinal ganglion ; Sp.c , spinal cord ; So, .somite ; N, notochord ; C.G., 

 ciliated groove of ventral surface of spinal cord ; Ac, gut. 



transverse section through the trunk of a very young Vertebrate 

 embryo, say a lizard ; and it is designed more especially to 

 show the nervous system. The neural tube {Sp. C.) is an open 

 plate, the two sides of which are folding together ; above it 

 are seen laterally two small outgrowths {Sp. Gl.) not connected 

 with the neural tube, and which have been split off from the 

 neighbouring skin or epiblast {Ep.) ; they are growing out from 

 the skin, and will soon be segmented off from it to form the 

 spinal ganglia. It has usually been supposed that the cranial 

 and spinal ganglia of Vertebrates arise as outgrowths of the 

 central nervous system. Such is not the case ; the diagram- 

 matic figure given above partially disproves that, and it would 

 be easy to give a series of figures which would demonstrate 

 beyond doubt that the spinal ganglia and certain portions of the 

 more complicated cranial ganglia arise from the epiblast outside 

 and beyond the limit of the central nervous system, and that not 

 a single cell of either cranial or spinal ganglia is derived from 

 the latter. Now, the mode of development of the ganglia in 

 Vertebrates tallies exactly with that described by Kleinenberg in 

 Annelida for the parapodial ganglia. These latter also arise as 

 epiijlastic differentiations just above the lateral limit of the 

 ventral cord, and, like the ganglia of Vertebrates, they appear 

 segmentally. 



Here, then, is one point of close resemblance, and not an 

 unimportant one, between the Annelid and the Vertebrate. 



Having got thus far, one is tempted to study the development 

 of the central nervous system of Vertebrates more closely, in 

 order to see whether other Annelid peculiarities recently dis- 

 covered are represented. Of such, two have presented them- 



selves, and while probably but the forerunners of a series yet to 

 be unravelled, they are in themselves of the highest significance, 

 I have represented these in the diagram in such a form that 

 they may be easily understood ; but be it remarked that they 

 do not appear so obviously till at a later stage than that 

 depicted. 



Evidence has long been wanting of a bilateral origin of the. 

 central nervous system of Vertebrates (no doubt it is a bilateral 

 structure — everyone, except perhaps Prof. Hubrecht, believes 

 that), and if it is comparable to tl-e ventral chain of Annelids, 

 it ought to show traces of such origin in its early development. 

 It is precisely this which I believe to have discovered. 



Even before the actual closure of the two limbs of the neural 

 plate occurs (a phenomenon which takes place much later than 

 is generally supposed) the neuro-epithelium of the one limb does 

 not pass directly into that of the other limb, for the two are 

 separated below by a tract of non-nervous epiblast having the 

 characters of a ciliated epithelium {C.G.). Thus, with greater 

 truth than one can speak of the absence of primary connection 

 between ganglia and central organ, it must be admitted that the 

 two lateral halves of the central nervous .system itself are at first 

 destitute of nervous connection with each other. This ciliated 

 groove (C. (7.) is a veiy obvious structure in transverse sections of 

 great numbers of Vertebrate embryos. Curiously enough, I 

 cannot find it figured by any embryologist except His, and he 

 does not say anything about it. 



The peculiarities just described (a developing nervous appa- 

 ratus composed of two bands of neuro-epithelium separated from 

 each other by a ciliated groove) are eminently characteristic of 

 Annelids. Nay, more ; Kleinenberg states that the ciliated 

 groove takes its origin from a double row of cells in the mid- I 

 ventral or neural line, and I am of opinion that such is ^ 

 the origin of the ciliated neural groove of Vertebrates. If this 

 discovery of the double nature of the neural plate has the sig- 

 nificance which I claim for it, the generally accepted opinion as 

 to its primary structure must fall to the ground. The neural 

 plate is usually supposed to be composed of two layers — 

 an inner nervous one, and an outer ordinary non-nervous one; 

 the inner layer is supposed to give rise to the nerve-cells, &c., 

 while the outer epiblastic one, having unfortunately got shwt in 

 with the tube formation, has nothing left to it but to form the 

 ciliated canal. 



Both these conclusions are wrong. Years ago, Altmann 

 showed — and it has been confirmed scores of times — that it is just 

 those cells next the primary central canal which increase most, 

 and so form the antecedents of the ganglion-cells. The real truth 

 is, that the greater part of the epithelium lining the primary (as 

 opposed to the permanent) central canal is a neuro-epithelium, 

 for only such a one has the faculty of producing ganglion-cells 

 on its inner side.^ 



The epithelium of this ciliated groove having developed cilia, 

 undergoes no further differentiation for some time ; it is the only 

 fart of the primary central cylinder which is ciliated, and which 

 does not form ganglion elements, and hence it is the only part 

 which is not neuro-epithelium. It forms later, by the growth 

 and increase of its elements, most if not the whole of the ciliated 

 epithelium of the permanent central canal. ^ 



In one respect the ciliated groove of the Vertebrate differs 

 from that of Annelids — it gets invaginated along with the central 

 nervous system ; and I am not aware that any Annelid is known 

 in which the ciliated groove is removed from the outer surface 

 of the body, along with the ventral cords. 



Now these facts are very remarkable, and, taken in connec- 

 tion with other points previously mentioned — such as the 

 formation of the notochord and swimming-bladder, the lateral 

 sense-organs, the origin of the ganglia — they furnish us with a 

 combination of Vertebrate characters for which a parallel is to be 

 found in the Annelida, and in no other group. 



Further, we have in the nephridia of Vertebrates a series of 

 structures which, as Semper first showed, find their parallel in 

 Annelids. When one considers recent advances in our know- 

 ledge of the nephridia of Annelida (more especially those we 

 owe to Drs. Eisig and Ed. Meyer) in connection with the, as 

 yet, partially unpublished researches of Dr. Van Wijhe on 

 Vertebrates, the justice of Semper's and Balfour's renowned 

 comparison becomes more and more obvious. 



I have shown, in a former number of Nature (vol. xxxvii, 



' As I write this, there occurs to me a beautiful idea of Kleinenberg's, that 

 the ganglion-cells in the central organ which are perceptive of light have 

 that power in virtue of the fact that they were themselves once retinal 

 elements or parts of such elements. 



