120 THE entomologist's HECOKD. 



where it falls abruptly over the ninth and tenth segments. The ventral 

 processes, which act as prolegs, are absent on the prothorax, but very 

 strongly developed on the meso- and metathorax, they are also present 

 on the other abdominal segments, except the eighth and ninth. Those 

 on the tenth abdominal are especially useful. In crawling on a flat 

 surface the larva arches its body, using the processes on the thorax and 

 the tenth abdominal, not making much use of those intervening. 

 There are no markings on the dorsum of the larva, but beneath are 

 some quadrate and kite-shaped dark patches which I take to be the 

 ganglia. The skin is much spiculated and the larva is provided with 

 simple tubercles and setae. The prothorax carries three pairs of setse 

 on each side of the median line. There is a single seta above the 

 spiracle and a longer one below. On the mesothorax ii is very close to 

 i, but on the abdominal segments ii is behind i and rather low down, 

 it has a short seta, while i has a long seta. Tubercle iii is very small 

 above the spiracle, while iv below carries a very long seta. Tubercles 

 v and vi appear to be absent, but I think vii is present. There are 

 three strong setae on each of the thoracic processes. The width of the 

 head is O'SSmm. and the length of the larva 4-75mm. In the tenth 

 abdominal segment two rods, 0-15mm. in length, may be seen. Their 

 anterior ends are attached to muscles which draw the ends near 

 together or separate them, while the posterior ends are attached by 

 muscles to a frame-work which runs round the posterior margin of the 

 tenth abdominal segment. These movable rods are present in all the 

 larval stages, and I have seen them in Kepticulid larvae, but do not 

 understand their use. I cannot find anything in the larva that would 

 prevent its being congeneric with Nepticula. Except in size this larva 

 does not differ from that of Nejiticida acetosae, but I have not compared 

 them side by side. The larva, in the fourth instar, continues its mine 

 down the broom twig. The mine takes up the whole of the furrow and 

 is much raised. It appears brown where the larva lies but all behind 

 the larva is black. The larva does not confine itself to one furrow, but 

 often passes through a ridge and down the adjoining groove. This 

 part of the mine runs to a length of about 70mm., so that the whole 

 mine from entrance to exit exceeds 100mm. When about to leave the 

 mine the larva bites a narrow transverse opening through the cuticle 

 and forces its head and thorax out of the mine. After this, by move- 

 ment of the posterior segments, it completes the exit, The larva 

 emerges venter uppermost and, curving its body into a horseshoe shape, 

 falls from the mine. Those I had, on reaching the table cloth soon 

 began crawling away. Unlike many Nepticulid larvae they do not 

 spin any silk at all during their progress. Some of the larvfe I put in 

 a flower pot filled with light soil, they all went down very soon. One 

 was placed in soft paper, but on opening this several days afterwards 

 I found the larva curled up. It had made no attempt to spin and was 

 quite lively, so it wjis transferred to the pot and it followed the example 

 of the others. Three weeks later I turned the earth out of the pot and 

 passed it all through muslin without finding any traces of the larvee, 

 either alive or dead. I imagine they must have somehow or other 

 managed to escape. From this experience I imagine that the larva on 

 leaving the mine passes the rest of the winter without forming a 

 cocoon. In the spring it probably comes forth either to spin its cocoon 

 or first, as I believe, Warren stated or suggested, to mine the leaves of 



