I. TRÄGÅRDH, LEPIDOPTEROUS LEAF-MINERS. 13 



and far forward as in that genus, but marginally and farther 

 backwards. 



Finally, in the fullgrown Lithocolletis (fig. 17), we find 

 still another arrangement, there being 4 ocelli left, 3 of which 

 are close together anteriorly. All are placed ventrally and 

 it seems reasonable to suppose that this type is derived from 

 the primitive one (fig. 8) simply by the reduction of the two 

 dorsal ones of the six ocelli. 



Summary. 



This change in the arrangement of the ocelli and, in some 

 genera, considerable reduction in their number are obviously 

 adaptations to the special circumstances which meet the 

 leaf-miners and which differ from those of an ordinary lepi- 

 dopterous larva. 



As is well known, physiologists generally conclude from 

 the anatomy of the ocelli and from experiments with insects 

 that these are comparatively imperfect optic organs, with 

 the aid of which the insects can only discern the surroun- 

 dings at a very short distance, and that in insects with 

 compound eyes they ha ve only an accessory value, and seem 

 to be useful in the dark. 



T am, however, not aware of any experiments having 

 been made on lepidopterous larvae, and we do not possess 

 any exact data of their visual powers. 



From their behaviour it seems, however, evident that 

 their faculty of sight is very small and that they are only 

 able to notice anything at very close quarters. 



The modifications of the ocelli, be it their altered arrange- 

 ment or the reduction in their number, seems to have been 

 brought about by the following features characterizing the 

 leaf-miners. • 



Firstly, the pronounced dorsoventral flattening of their 

 head-capsule, which is one of the chief characteristics in their 

 adaptation, and which is due to the circumstance that the 

 leaf-tissue offers the least resistance to a flat head capsule. 



By the head becoming flattened, the »genae» viz. the 

 portion of the capsule where the ocelli are to be found, 

 practically disappeared, being only represen ted by the lateral 

 margin where the dorsal and the ventral surface meet. 



