38 Psyche [April 



axis of the body and all were slightly curved, so that the oblong 

 plate which they formed together was concave on the dorsal and 

 convex on the ventral surface. They gave off a bluish-green light 

 most intensely from their dorsal surfaces, as their ventral surfaces 

 appeared to be covered with a vague mass of reflecting tissue. 



The specimens were killed in 50 per cent, and then transferred 

 to 70 per cent, alcohol, which seemed to preserve them sufficiently 

 well for subsequent study. The junior author sectioned a number 

 of the specimens and stained them with Heidenhain's iron hsema- 

 toxylin. He also obtained excellent views of the structures as 

 whole mounts or partial dissections stained in alum cochineal. 



Fig. 1, PI. 3, represents one of the larvse in lateral view. The 

 head is very small, with minute eyes; the long body grows broader 

 posteriorly and terminates in a bulbous enlargement or vesicle, 

 separated by a distinctly constricted region from the more anterior 

 segments. If this constricted region is counted as a segment the 

 abdomen has the full number, eleven, of segments characteristic 

 of insects. Hudson's figure of the larva (PI. 8, Fig. 1) is certainly 

 incorrect, since it represents no less than nineteen post-thoracic 

 segments and the outline of the terminal bulbous segment is defect- 

 ive. It is probable that this region in our figure is too small, as 

 it was drawn from a preserved and perhaps somewhat shrunken 

 specimen. 



The peculiar glutinous web on which the larva moves about is 

 evidently secreted by a pair of voluminous spinning glands situated 

 in the anterior two thirds of the body. Their posterior tips are 

 represented at SG in Fig. 2, which is drawn from a dissection of 

 the slender hind-gut and the accompanying Malpighian tubules in 

 the posterior third of the body as seen from the ventral side. 

 These Malpighian tubules are four in number as in other Diptera 

 and all come off at the same level, run forward a short distance 

 and then turn back with more or less convolution, applying 

 themselves to the sides of the slender intestine. They gradually 

 diminish in diameter till they become very tenuous and straight 

 in the constricted region just anterior to the terminal vesicle and 

 are there closely applied to one another side by side on the ventral 

 surface of the rectum. Then they separate along the median line 

 into two pairs and again come together but increase two and 

 one-half to three times in diameter in the vesicle where they 



