230 THE CANADIAN ENTOMOLOGIST. 



disproportionate to its appetite, and the thin walls had finally collapsed. 

 At all times a voracious feeder, it at last finishes its career at the base of 

 the plant, often boring down and out through the tuberous roots. These 

 latter are said to be the most poisonous productions of our local flora, yet 

 marginidens flourish upon them to a surprising degree, becoming at 

 maturity fat, uncanny " grubs," we might almost say, from which it would be 

 little expected that such beautiful moths should ultimately result. Then, 

 too, there is the unfortunate, unnecessary habit of leaving the burrow and 

 pupating in the ground. There is a tinge of regret connected with this 

 statement, as it recalls the slip given me by the scores of larvae that had 

 been located after a great amount of patient searching. The harvest of 

 pupae, upon which it was anticipated a finger could be placed at the proper 

 time, was not to be mine ; in its stead, experience of the usual expensive 

 nature. 



A note of August 4 reads : " After examining dozens of plants that 

 had contained larvte, but which Iiad now left for pupation, one chrysalid, 

 evidenly stung, is found in a burrow, an irregular opening for the moth to 

 get out having been made. Nearly all seem to have eaten down through 

 the roots, going out at the lower end and apparently burrowing downward 

 still, as though impelled by a desire of making some celestial acquaint- 

 ances. No pupre being found in the vicinity of the roots or elsewhere, we 

 shall have to reserve our revenge for the brood of next year." 



On June 30, larvte were in the second stage from the last. The 

 colour is a light brown, with a decided pinkish tinge. The first four 

 abdominal segments are much the darkest by reason of the subdorsal and 

 substigmatal lines being here discontinued. The dorsal line is very evi- 

 dent on all segments ; these lines yellowish white ; tubercle arrangement 

 already very conspicuous and normal ; on all abdominal joints I. exceeds 

 II. rather more than it does later. Lateral tubercles darker and more 

 contrasting than the dorsal. Head, shield and plate pale, testaceous ; the 

 shield as wide as head, and edged with black at the side. Length 1.2 

 inches. Next to last stage : A marked gain in length is noted, the colour 

 has faded, head and shield are more shining and conspicuous. Length 

 1.5 inches ; duration of stage ten to twelve days. 



Last stage : All colour contrasts lost, and appearance anything but 

 prepossessing. The soiled translucent colouring, which darkens percepti- 

 bly on first four abdominal segments, seems due to the internal fluids. 



