K 



coMpnuND /:v/':s of artiihoi'ods. 2!»6 



Altlioujxli the Riilijcc't will 1)«- iiuich iiK.ru iiitclliiriMc uftcr 

 otiier tvj)es Iiiivo been duffuribeti, it will not ha tiltugether out ol' 

 pliice liere to mention brielly my view of tlie Bteps through which 

 the characteristic arrangement of parts in an ommatidiuin like 

 that of SeruJis might have been brought about. Suppose a 

 circular area of the skin to be divided into five zones, one lying 

 within tlie other. Mark the innermost circle as (1), and the next 

 zone as (2), and the middle as (3), the fourth as (4) and the outer- 

 most as (5). Suppose, further, this circular area of the Hat 

 ectodermal surface of tlie body to sink down as a conical pit 

 with its open base turned towards the exterior. The cell or 

 cells lying in the innermost circle (1) in the middle of the area 

 will sink deepest, while the outermost circle (5) will retain its 

 original level. The cells lying in the innermost circle (1) are 

 the "hyaline cells"; those lying within the second zone (2) are 

 the retinulae; those lying in (3) are the vitrellae; those within 

 (4) are the corneagen, and those within (5) are the group of 

 pigmented cells which surround the dioptric portion of the 

 ommatidium. 



All that is necessary to convert this diagram into the omma- 

 tidium of Serolis is to reduce the lumen of the pit until it finally 

 disappears and the cells facing the lumen of the tube come, into 

 contact with each other, or else remain separated by the chitinous 

 structure; a faint line like that shown by x in P^igs. 1, 3 and 4 

 being all that remains of the axial cavitv, like that shown in 

 Fig. 2. 



The compound eye of Serolis is in this view nothing but a 

 collection of these depressions in the skin, which by virtue of 

 their aggregation attain tliu morphological and physiological 

 value of an organ. 



h. Talorchestia. 

 (Figs. 3, 3o, PI. XXIX.) 

 A glance at the figures will at once show the fundamental 

 similarity of the ommatidium of this Amphipod and that of 

 Serolis. Fig. 3, PI. XXIX, shows a single ommatidium of 

 Talorchestia. The outermost covering c is the corneal facet. 

 The two cells which lie beneath it are tlie corneagen {c. y). 

 These cells are not so conspicuous in the adult as in the young 



