1895.] NATURAL SCIENCES OF PHILADELPHIA. 165 



alternate, in vertical sections of the retina also, and with the utmost 

 regularity with the single central rod elements. Such alternating rods 

 and double cones are met with only in sectious cut through the 

 greatest diameter of the eye and exactly vertical to the maximum 

 thickness of the retina, as should be expected from the manner of 

 arrangement of the rods and cones, as seen in sections that are cut 

 horizontally through the bacillary layer and parallel to a plane 

 tangent to its convex surface. 



Max Schultze has figured portions of the bacillary layer of the 

 retina of mammals, as seen from the surface, and one sees in such 

 figures that the cones are repeated at regular intervals, the more 

 numerous rods surrounding the cones. In the retina of the salmon 

 this arrangement is reversed, or at least is not so obvious unless one 

 assumes that every double cone in this form is supposed to be 

 encircled by eight rods. In the salmon the cone elements are most 

 numerous; in mammals they are less numerous than the rods over a 

 very large portion of the retinal surface. The$e facts together with 

 the rest of what has preceded do not conflict, however, with the view 

 that the groups of cells met with in the retina of the young salmon 

 are fairly to be regarded as having arisen from a condition in which 

 such groups of rods and cones were more or less distinct and separate 

 as in the compound eyes of Arthropods. On morphological grounds 

 alone, it seems to me, that we are driven to some such conclusion. 

 No matter what convictions one may hold as to the utter untenability 

 of any hypothesis that seeks to derive the vertebrates from the Arthro- 

 pods or vice versa, it seems probable from the evidence now at hand, 

 that highly complex and specialized sense-organs of all kinds, includ- 

 ing, possibly, the retina of vertebrates; have arisen by the coalescence 

 of simpler, smaller and more numerous sensory organs. That the 

 sense-organs of both vertebrates and Arthropods have pursued 

 parallel paths in the course of their evolution seems highly probable. 

 The frequency with which specialization, reduction and even appar- 

 ent simplification, in some respects, of organs, in the course of 

 organic evolution, is associated with a process of coalescence of more 

 numerous structures having the same function, is very strongly in 

 favor of such a conclusion. 



The single groups of retinal elements in the retina of Salmo, com- 

 prising, as we have seen above, thirteen cells, I propose to call 

 retinidia in contradistinction to what is met with in the Arthropod 



