14 ARKIV FÖR ZOOLOGI. BAND 8. NIO 9. 



If, under these circumstances, the arrangement of the 

 ocelli were to remain unchanged the result of this would 

 inevitably be t hat one part of the ocelli would become 

 dorsal, another ventral. 



But this would mean that they would become covered 

 by the epidermis and the tissue of the leaf pressing closely 

 against the dorsal and ventral surface of the head, and in 

 all probability preventing the ocelli to serve in these direc- 

 tions, the necessity of which ability would also seem very 

 problematic. 



On the contrary it seems reasonable to suppose 

 that the only direction in which there is any need 

 for them to be able to discern things is the only one 

 in which they are moving i. e. in a horizontal plane. 



From this point of view it is easy to understand that 

 the only solution of the problem, how to make use of the 

 ocelli under the altered circumstances, is that they become 

 as nearly marginal as possible and we have above traced, 

 how in the different genera this change has taken place, the 

 extreme in this direction being reached by Tischeria amongst 

 the tissue-feeders. 



There are, however, as above mentioned, amongst the 

 tissue-feeders three genera which differ from the rest as yet 

 examined, viz. Lyonetia and Nepticula and Eriocrania. In 

 the former the ocelli seem to be adapted to seeing indepen- 

 dently forward and sideways, a necessity the explanation of 

 which is not easy to find. 



As to Nepticula, the great size of its single ocellus pro- 

 bably compensates for the small number, but it is also charac- 

 terized by its more backward position than in other genera 

 with single eyes, which implies its function being to see 

 more sideways. 



It is vefy suggestive that these two genera are charac- 

 terized by their very narrow mines, winding and twisting in 

 all directions. 



As to Eriocrania, it is a blotch miner and it is very 

 curious that, to judge from the position of the ocellus, it 

 should be able to see only forward. 



The ventral position of the ocelli of the fullgrown Litho- 

 colletis, on the other hand, is easily explained by its mode 

 of life. When the blotch mine has attained its definite 



