678 William Patten 



that one of the planes of di vision is more distinct than the others, aud 

 that the inner ends of the cells show traces of a paired arrangement. 

 We interpret these faets as indicating that thepresent, quadruple 

 retinophorae were originally double; and in fact we do find 

 certain Crustacea, the Amphipods and Isopods, in which the paired 

 arrangement is a permanent condition. The majority of these simpler 

 eyes formflattened, or slightly protuberant, surfaces containing butfew 

 ommatidia. An eye with the following features could, then, be con- 

 sidered as one of the least specialized and primitive forms of the com- 

 pound eye : a facetless cornea ; a thick corneal hypodermis ; a small 

 number of short ommatidia containing double retinophorae provided with 

 two, large, axial rods which form a nearly terminal crystalline cone ; 

 Semper's nuclei may be above or beneath [Musca, Hickson) the retinidia ; 

 and finally, the entire ommateum forms a nearly fìat surface. How, then, 

 does such an eye compare with the lateral ones of Spiders ? It seems 

 to me that the resemblances are so striking that they hardly require 

 enumeration. In Spiders, the terminal rods bave become 

 axial; forming a double refraetive body homologous with 

 the crystalline cone; in some cases, one or two (?) nuclei bave 

 shifted their position so as to lie above the rods. That the apical posi- 

 tion of the nuclei, above the rods, in the compound eye, presents no 

 morphological difficulty, is shown by the frequeucy with which their 

 position is changed in Spiders, sometimes being above, and some- 

 times below the rods. In the simplest ocelli with terminal rods, the re- 

 tinulae form a single, or probably, at most, a double zone; with the 

 formation of an ommateum by the development of axial rods, the 

 retinulae show a strong tendenc}^ to form two or three zones sur- 

 rounding corresponding parts of the retinophorae. Grenacher has 

 net described any nuclei in the retina of Spiders, except those of 

 the retinophorae ; but we may be sure that the pigment is deposited 

 in distinct cells, just as in Scorpions and Limulus ^ whose nuclei 

 bave been overlooked by Grenacher and misiuterpreted by Graber. 

 In Spiders, as has been shown by Lankaster and Bourne for Scorpio 

 and Limulus, the outer and middle nuclei of Graber must be referred 

 to an outer and middle (and inner? row of pigment cells, or retinulae! 

 Then each ommatidium of the middle eye of Spiders would consist 

 of a double retinophora with axial rods surrounded by two or three 

 circles of retinulae (PI. 32, figs. 137 and 146). Such an ommatidium 

 agrees, in ali essential respects, with an ommatidium of a compound 

 eye like that of Gammarus. The only change necessary to convert 





