228 THE CANADIAN ENTOMOLOGIST. 



triangle. On the abdominal segments, tubercles I. and II. are placed in 

 the usual oblique setting, with III., III. a, IV. and V. clustered about 

 the spiracle. The eighth abdominal segment has I. and II. very large 

 parallel to the dorsal line, and with their opposites form the corners to a 

 conspicuous square. Preceding the anal plate on last segment is a lesser 

 plate or tubercle that occasionally becomes merged into the borders of 

 the former. 



On the seventh abdominal segment, IV. is sometimes raised to a little 

 above the line of the spiracles instead of lower down as is common to 

 Noctuids generally. 



Dr. Dyar has pointed out this feature as occurring in purpurifascia, 

 and in theorizing on the unusual break, concludes it has been an acquisi- 

 tion — or, rather, a transposition — to more fully protect the spiracle. (Jour. 

 N. Y. Soc, VII. 70.) A curious and perhaps significant fact is that this 

 break occurs in the root-feeders alone, at least so far as observations have 

 been made. Such larv«, burrowing down to the extremities of roots, are 

 at all times cramped for room, and are rarely able to turn round in their 

 galleries when desiring to go in a reverse direction. 



Can it be that this habit, which necessitates the constant backing of 

 the larva to the ground orifice for the disposal of frass, and which practi- 

 cally means as much backward as forward movement, is accountable for 

 this ? With the stem borers there is generally an extended burrow of 

 ample diameter ; if they wish to go in a reverse direction there is room 

 for turning, and we may assume a forward motion predominates. 

 Certainly the excess of backward movement made by the former would 

 bring an undue amount of friction behind the spiracle on joint seven, and 

 tubercle IV. has been raised to the point of greatest efficiency, the same 

 as it is on the other segments. This is merely a random idea, of thin air 

 consistency perhaps, and is advanced only that attention may be directed 

 here more fully. 



In dealing further with larval characteristics, it may be noted that 

 there is a tendency on thoracic segments two and three for the skin to 

 assume a puckered or roughened aspect, most notable in immature 

 stages, especially when the head is retracted. Inflated examples show this 

 invariably when but little air-pressure has been exerted in drying. 



The extensile, glandular process that exudes from the under side of 

 joint one does not seem to have been mentioned previously. It consists 



