272 [November, 1893. 



head about one-sixth the diameter o£ the field, that of the second head 

 about one-fourth, of the third about one-half, and of the fourth and 

 last as nearly as possible a full diameter. Hence, with the aid of this 

 scale, the existence of an early and otherwise unsuspected moult may 

 be almost as conclusively proved as if the tiny head were actually in 

 the field before one ; and since I have in this way been able to infer, 

 if not quite to prove, the existence of three moults in every species I 

 have examined, we may fairly conclude that three is the usual, if not 

 the invariable, number in a Nepticula larva. 



As regards the segmentation of the mine, the ease and certainty 

 with which it can be effected varies considerably, and in some cases it 

 must be left alone altogether. The relative breadth of the parts helps 

 but little, for the mines widen insensibly, or at any rate, without any 

 marked emphasis at the moulting spots, so that we are driven entirely to 

 the contents — the frass, for our characters. In the portion answering 

 to the first and second skins of the larva it is not possible as a rule to 

 resolve this substance into its separate grains, for it looks as if, from 

 having been deposited in a soft condition, it had run together into a 

 homogeneous cake. In the second portion (answering to the third 

 skin of the larva) the frass sometimes continues to retain this con- 

 cretionary character, but usually it is more or less completely grained, 

 and offers nearly as much variety in the arrangement as it does in 

 the third or final portion (answering to the 4ith skin), yet, as it by 

 no means follows that the same arrangement is adopted in both, 

 a ready means for their distinction is afforded. Occasionally, too, 

 a change in the colour of the frass will mark the spot where one 

 portion ends and another begins. But it will be better, perhaps, 

 if I give an example or two of the practicability of this segmen- 

 tation. The long gallery of the birch-feeding lapponica is well 

 known. In its first length the frass is of the typical concretionary 

 character, and completely fills the mine ; in the second length it is 

 coiled (form 3), but rather obscurely, from a tendency the grains 

 still have to run together ; in the third length the frass, which had 

 hitherto filled the mine, is suddenly collected into the centre, leaving 

 a wide margin on each side (form 1), and at the same time its colour, 

 which up to this point had been green, changes to black. Nothing 

 can be more satisfactory or precise than the mapping out here. Let 

 us now look dXfulgens. First comes, as usual, the concretionary be- 

 ginning, embracing the first two larval skins, then follows the second 

 length, characterized by unmistakeable coiling (form 3), and lastly the 

 third length, with the coiling gone and the frass scattered irregularly 



