73 



laid being milk white, with the shell extremely thin and pliable. A very fine net- 

 like sculpture is barely visible even under a strong magnifying gl 



The larva. — The more obvious characters by which the full-grown larva | fig. 3, b) 

 may be distinguished are as follows: 



Body depressed (but not flattened ) , elongate, very little tapering posteriorly, gla- 

 brous, color yellowish white, the head, larger portions of first thoracic segment, 

 upper side of anal segment, and the legs brownish or blackish. Head subquadrate, 

 about half as wide as the first thoracic segment, brown, shining, impunctate, with 

 a deeply impressed median line. First thoracic segment more than twice as wide 

 as long, a little longer than the head, distinctly Insinuate at the anterior margin, 

 sides strongly rounded; a large brown spot divided by a narrow median line occupies 

 the greater portion of the upper surface, but does not reach the posterior and side 

 margins. Second thoracic segment a little shorter, but wider than the first, .-ides 

 very strongly arched but not gibbous, surface uniformly whitish; third thoraci' 

 ment equal to the second. First abdominal segment a little shorter than the last 

 thoracic one and also a little narrower, but produced on each side into a triangular, 

 semitransparent tubercle, capped with a more horny point which in its turn termi- 

 nates in a short spine. Second, third, and fourth abdominal segments equal to the 

 first; on the following segments (which are slightly longer than the first J the lateral 

 tubercle is directed gradually more backward, the anal segment is unarmed, shorter 

 and much narrower than the preceding, subtruncate at tip. its upper side being of a 

 brown color. 



The general color of the dorsal surface is not uniform, but variegated by the trans- 

 parency of the skin as the fat corpuscles of the body appear on the surface as yellowish- 

 white spots, the rest of the body being grayish white. The sculpture of the dorsal 

 surface (excepting head and anal segments, which are smooth) consists of a very reg- 

 ular fine granulation and besides this of vague impressions arranged as follows: < >n 

 the first thoracic segment an undulating impression each side on the disk; on the 

 two following thoracic segments a long transverse median impression accompanied 

 each side posteriorly by a foveiform impression; on the abdominal segments a trans- 

 verse median impression and another oblique one each side, causing the spiracles to 

 be placed upon a kind of blunt tubercle. 



The ventral surface is colored like the dorsal and only the head and median por- 

 tion of the first thoracic segment is brown, and the general sculpture is like that of 

 the upper side, legs being brownish or blackish, stout, and widely separated. 



There are nine pairs of rather conspicuous stigmata, all visible from above and sit- 

 uated as follows: One pair at the anterior angle of the second thoracic segment, one 

 pair on each of the first seven abdominal segments, each stigma being at the base of 

 the lateral tubercle a little before the middle of the segment; the ninth pair is situ- 

 ated on the dorsal side of the anal segment. The thoracic' and anal stigmata are 

 larger than the intermediate ones." 



«It is interesting to observe that most leaf-mining larvae of different families or 

 even of different orders of insects exhibit a strong uniformity in general appear- 

 ance, viz, a more or less depressed body, the head being much narrower than the 

 first thoracic joint, while all joints of the body (excepting head and anal joint) are 

 strongly arched at the sides and often tubereulate or even spinose. Thus the larvae 

 of our leaf-mining Buprestidse (genera Brachys and Trachyscelus) and Khvneho- 

 phora (genus Orchestes) have an unmistakable resemblance to our Hispine Lame, 

 and all of these agree in shape with the numerous Tineid leaf-miners. 



There can be no mistake about the number and position of the stigmata, and the 

 account given by Chapuia & Cande/.e appears to be based upon a wrong interpreta- 

 tion of Harris's figures, entirely ignoring the correct description i)i that author. 



