634 William Patten 



form one, as necessarily should be the case provided these figures 

 are formed by the ends of seven retinulae. There is, however, one 

 other supposition possible, and that is that the median, scalloped re- 

 tinula [1] is really a double cell, a not improbable condition wheu we 

 consider its large size, its median position compared with the paired 

 arrangement of the two opposite ones, and its peculiar shape, which 

 probably is connected with the performance of some special function 

 differing from that of the other retinulae, This supposition would be 

 changed almost to a certainty, if it could be shown that for each om- 

 matidium, there were eight Segments in the tubes instead of seven. But 

 as I have Said the difficulties in the way of accurate Observation were 

 too great for the decision of this question. 



The seven retinulae surrounding the pedicel are divided into two 

 groups composed of four black and three brown cells. The median one 

 of the three latter is remarkable on account of its greater size, and pe- 

 culiar shape. At the beginning of the laminated structure of the pedicel, 

 the axial wall of this cell becomes scalloped, each fold projecting into 

 the end of a primary piate (fig. 72). I can form no conjecture as to the 

 meaning of this remarkable structure. If my memory does not fail me, 

 the same condition prevails in Galathea. The structure was then un- 

 known to me, and I failed to give sufficient attention to, or to make 

 a note of it. 



The pigmented c oliar of the retinophorae is formed by a third 

 circle of four cells arranged in two pairs, the darker ones forming 

 the inner, and the lighter ones the outer circle of the collar. Each one 

 of the inner cells is triangulär in shape, the apex being directed out- 

 wards, and the thickened base inwards. Both halves of the cell are 

 likewise triangulär, but lie in planes at right angles to each other, the 

 line of union being thickened to form a deeply pigmented ridge (fig. 69), 

 smallest anteriorly, and increasing in size posteriorly, until it is con- 

 tinued beyond the base of the cell as a pigmented rod (fig. 78), which, 

 after losing its pigment, is continued still farther inward, to the basal 

 membrane, as a slender, colorless rod, or bacillus (figs. 79 — 100, bc). 

 In cross sections of the collar, the halves of the pigment cells are seen 

 to diminish in thickness at their edges, so that at the intermediate Cor- 

 ners there is hardly any pigment at all ; but the calyces are so closely 

 packed that the thickest part of one pigment cell lies against the thinnest 

 part of the other, the result being that each calyx seems to be sur- 

 rounded by a uniformly thick band of pigment, which is better devel- 

 oped toward the inner edge of the collar than in the opposite direction. 



i 



