Campbell. — Notes on the Ble pha rocericlae of New Zealand. 259 



My own observations on these larvae are that all three have a dorsal 

 armature with special tubercles corresponding in number and position 

 on their respective segments — i.e., 12 cephalothoracic, 14 on each body- 

 segment, and 18 on the double 6th segment. The primary spines on A are 

 large, sharp, and black in colour ; those on B and C are transparent and 

 cone-shaped. The larva B is larger than the others, as it should be if it is. 

 as I suspect, related to N. kudsoni. 



Bezzi describes the gills of the larvae as follows : Larva A — " Tracheal 

 gills in single series, those of the anal clump distinct " (1914, p. 118). 

 Mr. D. Miller, who has translated Bezzi's useful paper,* after examining 

 Dr. Chilton's specimens says, " Not distinct, apparently six in number." 

 Larva B — " Tracheal gills forming a small indistinct tuft near the anterior 

 margin of the segment, those of the anal tuft apparently distinct" (p. 118). 

 Larva C — " Suckers large, while the tracheal gills placed on the anterior 

 margin of the segment are small but distinctly visible " (p. 124). 



Bezzi was not quite certain about the gills, and he makes no note of other 

 important characters ; but this can be understood, as his specimens appear 

 to have dried up. His description of the tracheal gills of Curupira is : 

 " Tracheal gills not arranged in tufts, but forming 2 rows running from 

 the anterior to the posterior margin of the segment, one on the right, the 

 other on the left of the sucker, consisting of from 6 to 8 gills in each row ; 

 anal tuft composed of 4 branches, antennae very short, 2-jointed, lateral 

 processes simple but very short, dorsum bearing powerful spines." (Bezzi, 

 1913, p. 76.) In C. torrentium there are 8 tracheal gills in each row. 



Excepting the number of dorsal spines, this arrangement agrees fairly 

 closely with that found in larva A. Moreover, a comparison of figs. 1 to 27 

 will show that in the form of the dorsal spines and the arrangement of the 



/*//-> z 



Fig. A. — 1. Wing of Curupira torrentium (F. Mull.), (after Bezzi). 

 2. Wing of Apistomyia elegans (Big.), ? . 



tracheal gills larva A stands apart from larvae B and C, which resemble 

 each other in dorsal armature, but differ in the form of the lateral processes 

 (figs. 38-40) and in the number and arrangement of the gills. Larva A 

 apparently belongs to the Paltostominae, and larvae B and C to the 

 Apistomyinae. 



To turn now to the imagines : Bezzi gives a diagram of the wing of 

 Curupira torrentium F. Mull, (see fig. A, 1, of this article) which shows this 



* Mr. Miller's manuscript translation is deposited in the Dominion Museum, 

 Wellington. 



9* 



