54- 



NA TURE 



[Eeb. 12, 1 88; 



they are always pigmented either by a diffuse fluid " re- 

 tinal purple,'' or by pigment in granules, or both. 



In both these particulars the retinulae of Arthropoda 

 resemble the nerve end cells of other animals. 



It is hardly necessary to point out that Leydig, Max 

 Schultz, Grenadier, and many others, have traced the 

 optic nerve fibrils to the retinulae. I have in my posses- 

 sion several series of preparations showing this both in 

 insects and Crustacea, and any one can readily see this 

 for himself by making even clumsy sections through the 

 eye of Squilla. 



In Fig. i I have figured the nerve fibrils of the eye of 

 Squilla perforating the basement membrane and entering 

 the retinulse, and in Fig. 2 a transverse section through 

 the rhabdom and retinulse showing their relat ve position 

 and numbers. 



A special feature of the retinula is that it is always pig- 

 mented. In specimens hardened in spirit a granular 

 pigment may be seen in the retinula cells, which is 

 usually of a light-brownish colour and very unevenly dis- 

 tributed (Fig. 2). But in addition to this granular pig- 

 ment, the retinulae contain a true retina purple, which 

 fades upon exposure to the light. This was discovered 

 in 1864 by Leydig 1 in the following genera of Insecta : — 

 Procrustes, Scarabaeus, and Pieris, and in Astacus among 

 the Crustacea. I have also seen it in Afusca vomitoria, 

 and have now no doubt that it exists in the Arthropod 

 eye generally. 



So far, then, I think it must be admitted that both 

 anatomical and physiological considerations tend to 

 prove that the retinula is the nerve-end cell of the 

 Arthropod eye. 



When we turn to morphology, too, we have confirma- 

 tory evidence that this is the case. 



In the ocellus of the water-beetle larva the retina is a 

 simple cup of pigmented hypodermis cells, in which the 

 optic nerve fibrils may be readily seen to terminate. 

 These cells are most certainly homologous with the 

 retinula cells of the so-called " compound " Arthropod 

 eye, as has been already shown by Grenadier in his im- 

 portant treatise, " I ntersuchungen uber das Sehorgan 

 der Arthropoden " and elsewhere, and confirmed by the 

 more recent researches of Lankester and Bourne upon 

 the eyes of Limulus and the scorpions. 



The researches of Claparede and Weismann on the 

 development of the eye of Arthropods confirms the 

 deductions of morphology, by proving that the cells which 

 ultimately form the retinulae are specially modified 

 hypodermis cells, and at an early stage come into con- 

 nection with optic nerve fibrils. It any further evidence 

 were required to confirm this homology it can be readilv 

 obtained by studying the eyes of very young cockroaches, 

 in which the retinulae at the periphery of the eye are 

 formed from specially modified and deeply pigmented 

 fi . podermis cells. 



But it is tedious and unnecessary bringing evidence of 

 this kind to confirm a theory which is already fullv esta- 

 blished in the minds of most naturalists. In fact we have 

 here an instance in which morphology, physiology, com- 

 parative anatomy, and development combine to establish 

 an homology, and consequently we must definitely assert 

 that the retinulae are the nerve-end cells throughout the 

 Arthropoda. But what is the meaning of Lowne's bai ill ir 

 layer behind the basilar membrane ? and does it exist in 

 all Arthropods? 



It- is perfectly true that behind the basilar in many 

 Diptera, Coleoptera, Lepidoptera, and Hymenoptera there- 

 is a layer composed of a number of small cylindrical 

 masses which has a superficial resemblance to the rods of 

 the Vertebrate eye, but Mr. Lowne did not discover this 

 layer in any sense of the word, for it was perfectly well 

 known to Leydig, who figured it in Formica ru/a, Dytiscus 

 all's, and Sphinx ligustri (vide Leydig's Tafeln, 



" Tubingen, 1S64. 



viii. ix. x.). The little cylindrical masses cannot be 

 regarded as cells, nor rods, nor bacilli, for each one of 

 them is composed of a very fine reticulum of nerve fibrillar 

 which is in direct communication with the optic nerve 

 fibrils behind, and the terminal anastomosis of the optic 

 nerve fibres in front. In fact, these "bacilli " of Lowne are 

 connected with nerve fibrils on both sides, and thus differ 

 from " nerve-end cells " in one of their two fundamental 

 characters. 



Very often, too, this layer is quite devoid of any pig- 

 ment (Apis, Eristalis, Bombyx, Squilla, &c), and no one 

 has ever yet been able to demonstrate the presence of 

 retina purple in this region. 



Another important difficulty in the way of accepting 

 this theory, too, is the fact that this layer is not always 

 present (Periplaneta, Nepa), and in all Crustacea and 

 many insects it cannot be divided into separate bacilli. 



I have lately paid considerable attention to this part of 

 the optic tract, but must defer a fuller explanation of the 

 meaning of it until I am able to publish my paper in the 

 Quarterly J oitmal of Microscopical Science, when I shall 

 be able to illustrate my researches by several figures. To 

 summarise, however, the evidence against this layer being 

 composed of nerve-end cells : We find that it is certainly 

 not homologous with the retina of other animals : optic 

 nerve fibrils both enter and leave it ; it is devoid of retina 

 purple or of any other form of pigment in many Arthro- 

 pods, and finally it is absent as a bacillar layer in many 

 insects and in all Crustacea. In fact we can bring as 

 much evidence to prove that this is not the retina as we 

 can to prove that the retinula; are the true nerve-end 

 cells. 



At the conclusion of his paper Mr. Lowne says, in re- 

 ferring to a recent memoir of Justus Carrie-re of Strass- 

 burg, " He remains, however, a disciple of established 

 views, and has not given the retinal layer nearly so much 

 attention as it deserves.'" I have given the retinal layer 

 as "much attention as it deserves," and must also claim 

 to remain a "disciple of established views." 



Sydney J. Hickson 



RORAIMA 

 A TELEGRAM has been received at Kew giving the 

 •**- welcome news that Mr. Everard F. im Thurn has 

 at last ascended Roraima. This has been the cherished 

 object of botanical exploration in South America for the 

 last quarter of a century. The expenses of Mr. im 

 ["hum's expedition have been borne in equal shares by 

 the Government grant of the Royal Society and the 

 Royal Geographical Society. 



The latest news from Mr. im Thurn was in a letter 

 dated December 6 from the south side of the moun- 

 tain, and the following passage describes the position 

 immediately before the final attack : — 



" Before we came to Roraima itself we had four days 

 walking through a purely savannah, but most glorious 

 country, and over splendid mountain passes, guided by 

 an Arecoona who said, villain that he is, that he knew the 

 \\a\ to Roraima. But at a village marked on the map as 

 Ipelemonta, on the Aroopa River, and with a consider- 

 able mountain pass still between us and Roraima, our 

 villain guide at last admitted that the road for some 

 distance had been quite new to him, and that he now 



ii n not how to proceed further. However, at last we 

 procured a guide, and came, in some four hours, out of 

 our difficulties at Ipelemonta (its real name, by the way, 

 is Toorarking), into this inconceivably magnificent valley, 

 and are installed in a village on the actual southern slopes 

 of Roraima itself. 



Yesterday Perkins and I ascended the slope of Roraima 

 to a height of 5600 feet to a most beautiful spot — a very 

 garden of orchids and most beautiful and strange plants. 

 To-morrow, after despatching the bearer of this scrawl, we 



