Eyes of Mollusca and Arthropods. 68 1 



ommatidia. Tbis is tlierefore just the opposite to what takes place 

 in the higher Arthropods, where a compound eye is due to the modifi- 

 cation of a single ocellus. I am conscious that the opponents of the 

 latter view might see here a contradiction, and could urge with some 

 plausibility that, if the lateral eye of Lwiulus is due to an aggregation 

 of many ocelli, it furnishes just the Illustration desired to show that the 

 faceted eye was produced by a similar, but more extensive series of 

 changes. But we have already given reasons for believing that the 

 changes undergone by the eyes of Limulus would never lead up to a 

 typieal compound eye, — in fact their development has been in a dia- 

 metrically opposite direction — and that the resemblance is of the 

 most superficial and iusiguificant nature. 



The condition represented by the eyes of Limulus and Scorpio is 

 as far removed from the primitive type, as are the compound ones of 

 Insects and Crustacea. But this condition is attained by a great reduc- 

 tion in the number of ommatidia without a corresponding increase in 

 their functional powers. When \nq compare the direction of their 

 development with that of the ocelli which give rise to the compound 

 eyes. we must admit that, if the development of the latter is upward, 

 towards organs of greater structural complexity and functional activity, 

 then the former are tending in the opposite direction, downwards, 

 towards greater structural simplicity and less perfect functional activity. 



In the Myriapods, the eyes have remaiued nearly stationary; 

 there has been no great change in the number of ocelli, or in the con- 

 dition of the ommatidia, which retain their primitive characteristics, in 

 that their rods are terminal and form a continuous layer, a retineum. 



In the Spiders, most of the ocelli have likewise remained nearly 

 stationary, the ommatidia forming a retineum; others, the posterior eyes, 

 have undergone important changes, not so much on account of their 

 extent, as direction. These changes carry such ocelli nearly 

 to the le vel of typieal compound eyes. The changes consist in the 

 development of double, axial rods. terminal nuclei for the retinophorae, 

 and two or three ;? circles of retinulae, an outer (middle?), and inner 

 one. The primitive, double retinophora is in direct contrast with the 5 

 and 10 fold retinophorae oi Scorpio a,nà Limulus. To change such an 

 ocellus into a compound eye, we have only to reduce the corneal lens 

 to a thin layev and flatten the ommateum. A necessary result of these 

 changes would be a more perfect Isolation of the retinophorae, and con- 

 sequently a better development of the retinulae. A compound eye being 

 once formed, it is further perfected by the development of a corneal 



