294 r. M. HOWLETT. 



The pattern shown by the slight sculpturing on the surface easily differentiates 

 the species, as follows :— 



Lines on the surface slightly wavy and irregular, joined at irregular 

 intervals by fainter cross-lines , . . . . . . . P. papatasii. 



Longitudinal lines more numerous and more regular, as are also the cross- 

 lines which join them, so that the pattern appears to be made up mostly 

 of narrow parallelograms . . . . . . . . . . P. argentipes. 



The pattern appears to be made up of juxtaposed circles forming cells of 

 roughly hexagonal shape . . . . . . . . . . P. minutus. 



The egg-patterns are shown on PL xi., figs. 1, 2, 3, and 5. 



Characters of newly-hatched larvae. 



The two characters which are most quickly and easily recognised are the relative 

 length of the tail-bristles as compared with the head and body, and the shape of the 

 body-bristles. The dorsal bristles of the segments forming the posterior part of 

 the abdomen are generally easiest to examine, and the three characteristic shapes 

 are best seen in the figures on PI. xii. These figures were verj^ carefully drawn with 

 the camera, and show the bristles in newly hatched and adult larvae. 



The points of difference are :• — 



(1) The straightness or curvature of the bristle. 



(2) The relative thickness of the apical and basal portion compared with its 



total length. 



(3) The outline of the apical portion. 



The relative length of the large tail-bristles as compared with the body can be 

 roughly expressed thus :■ — 



Ratio of length of body with head to length of tail-bristles = 6:5 in papatasii 

 larvae, = 5 : 5 in argentipes, = 4 : 5 in minutus. 



Hatching of the Egg. 



Just before hatching a slight motion of the egg can sometimes be detected, 

 ultimately resulting in the appearance of a longitudinal split in the antero-lateral 

 region, extending about half the length of the shell. Probably this split is started 

 by means of what we have called the " egg-tooth," but the shell is not sufficiently 

 transparent to enable one to make sure of this. Figs. 4 and 5, Plate xi., show 

 the young larva hatching. It gets free of the egg-shell in somewhat less than a 

 minute. There is often another split in the shell running round to the dorsal 

 surface. AVhen this second split is not made, the shell does not collapse after 

 hatching, but, save for its lighter colour, looks as if it were intact. The charac- 

 teristic pattern can be well seen on the surface of empty egg-shells, and this, 

 combined with the relative length of the body and tail-bristles of the newly 

 hatched larva, gives an extremely easy and reliable identification of any one of the 

 three species mentioned in this paper. 



The habits of the larvae are interesting, but cannot here be dealt with 

 at length. The function of the tail-bristles is not yet fully understood, 

 although they form such a characteristic feature. The larvae cannot burrow. 



