1. TRÄGÅRDH, METAMORPHOSTS AND BIOLOGY OF ORCHESTES. 9 



coalesces with the preceding one and in consequence all together 

 form a blotch. Or, the larva eats indiscriminately to the 

 left and the right, in which case the blotch arises in that 

 way, that concentrical streaks of the parenchyma are de- 

 voured successively. 



I have not had the opportnnity of watching these processes, 

 but it seems probable that the blotch-mine of O. popiili is 

 formed in the latter way. 



Once the blotch is large enough to allow the larva to 

 lie extended in its full length, it feeds in another way, 

 which I have been able to follow under the microscope; it 

 lies then at a right angle towards the perifery and eats small 

 portions out of the low walls formed by the parenchyma; 

 sometimes it makes a deep pocket at one place, but generally 

 it turns round, each time enlarging the mine by a narrow 

 streak. 



Consequently the larva is comparatively stationary, and 

 this fact obviously accounts for its being much less amply 

 provided with cuticular spinulse than the larvae of the two 

 other species. 



As Letzner has already pointed out, the larva does not 

 consume the parenchyma entirely, but only a comparati- 

 vely thin layer of it. Letzner sees in this an act of in- 

 telligence: the larva does so, because, if it devoured the 

 whole parenchyma between the upper and the lower cuticle, 

 the lower cuticle would become as thin as the upper one; 

 in consequence the larva would be exposed to the wind, sun 

 and rain and soon perish. 



But a more simple explanation is, that, as the mandibles 

 work in a horizontal plane, they are only able to detach 

 exactly as much as can be caught between the upper and 

 lower tooth, and as the leaf is comparatively thick, the rest 

 of the parenchyma remains intact. That this is the explana- 

 tion, is evident from a comparison with Gracilaria syringella, 

 which I had an oppurtunity of studying last summer. The 

 larvae of this moth have during the first two stages a flat, 

 quite horizontal head, and the thin mandibles work in a horizon- 

 tal plane. Consequently only a very thin layer of the paren- 

 chyma is consumed. But låter, when the galleries have be- 

 come larger, the larvae in the 3rd and 4th stages, which are 



