242 MYRIOPODA. 



the outer wall becomes thinner while the inner wall thickens. The 

 whole of the latter yields the retina, and the former secretes the lens. 

 According to this author this outer layer has not the significance of a 

 vitreous body-layer, but only functions to secrete the lens, and at 

 a later stage degenerates (Fig. 123, 7i). Between the regularly arranged 

 cells of the future retina, Heathcotb found smaller cells of irregular 

 form resembling amoeboid mesoderm-cells, and he assumes that these 

 wander in among the ectodermal retinal elements and function as 

 pigment-cells. A layer of mesodermal-cells becomes closely applied 

 to the retina, forming a capsule round the optic cup (Fig. 123, k). 



Heathcote's account of the rise of the Myriopodan eye does not sufficiently 

 explain either its manner of development, or the relation of the ontogenetic stages 

 to the adult eye. It is specially important to ascertain the significance of the 

 outer layer of cells lying beneath the lens (Fig. 123, h) in the adult eye, or at 

 least to prove its presence. A species of Julus examined by Gkexacher has a 

 unilaminar basin-shaped eye, devoid of the vitreous layer. Other Myriopoda, 

 especially Chilopoda as it appears, have the vitreous body, and thus have 

 bilaminar eyes. These eyes thus differ somewhat from ocelli of the simplest kind 

 such as occur in other Myriopoda. It would therefore be importaut to prove 

 whether the rudiment of the eye is actually a closed vesicle, or whether it is not 

 rather an ectodermal depression, such as it appears to be still in the adult in the 

 case of eyes devoid of the vitreous body. The vitreous body in such cases would 

 have to be explained as arising, as in the ocelli of the insect larvae (Fig. 36), 

 by the ingrowth of ectoderm from the sides. 



As we have already seen (p. 194), the eyes of Peripatus arise as ectodermal 

 depressions, and it seems the simplest explanation if we can trace back the 

 basin-shaped eyes of the Myriopoda as well as the ocelli of the Insecta to a 

 similar primitive method of development resembling that of the Annelidan 

 eyes (p. 69). The Myriopoda, indeed, are of special interest in this connection, 

 inasmuch as some of them possess only a few ocelli {Scolopendra, for example, 

 has only four on each side), while a larger number is found in others, leading 

 to the formation of so-called crowded ocelli (thirty to forty and more, in Lithobius, 

 Julus, Fig. 119, oe), till finally a compound eye results, consisting of numerous 

 single eyes (about 200 in Scutigera), and resembling a facet-eye like that of 

 Limulus. This may even show a kind of rhabdom-formation without belying 

 its relation to the crowded ocelli of other Myriopoda. AVe thus appear to have, 

 in the Myriopodan eyes, the different stages indicated by us as of probable 

 occurrence in the phylogeny of the facetted (polymeniscus) eye when we were 

 considering the Arachnidan eye (p. 69).* Careful ontogenetic and morphological 

 investigation of these points is exceedingly desirable, and would no doubt have 

 a successful issue. 



* [Recent investigations on the structure of the eyes of Scolopendra and 

 Scutigera (Nos. I., YIL, XI.) entirely confirm this suggestion. The eye of 

 Scolopendra consists of a few perfectly distinct ocelli, each of which exhibits 

 the structure typical of that class of e}*e, being a simple ectodermal depression 

 with lateral vitreous body and numerous retinulae, each with a rhabdom pro- 

 jecting into the cavity of the eye, the whole being overlaid by a cuticular lens. 

 In Scutigera, on the other hand, the ocelli are greatly modified by mutual 

 pressure until they have almost assumed the character of the omraatidia of a 



