534 MODERN CLASSIFICATION OF INSECTS. 
body moderately long (_fig.127.12. Subula maculata) ; the antenne are 
10-jointed *, without any apical style or seta (/fig.127.13. ant. of 
Subula), the eight terminal joints forming an elongated nearly cylin- 
drical mass ; the wings, when at rest, are laid upon the abdomen ; the 
tarsi have three pulvilli, both in Beris and Subula; the scutellum is 
spined in some species (Beris); the terminal veins of the wings are of 
the ordinary thickness, and are disposed in a manner intermediate be- 
tween those of the Stratiomide and Tabanide, to which last indeed 
Subula makes a close approach in this respect. 
The species are mostly found in damp situations, and are of slug- 
gish habits. 
M. Van Roser (Naturwiss. Abhandl. or Ann. Nat. Hist. Soc. Tubin- 
gen, vol. ii. 1828; Meig. Zw. Ins. vol. vi. p.319.; and Ferussac, Bull. 
Sci. Nat. June 1829) discovered the larva of Subula varia in an oak 
tree; it is of a brown colour, elongated, very much depressed, rather 
attenuated at each end; the body is composed of twelve segments, 
furnished on the upper side with hairs arising from small tubercles. 
The pupa is inactive, and is described as preserving the previous 
form of the insect. The real pupa, however, is enclosed within the 
skin of the larva, which does not change its form, thus agreeing with 
the other typical Notacantha. This is clearly shown in a specimen 
of the exuvize, both of the larva and pupa of Subula varia, given to 
me by M. Van Roser, and represented in fig. 127.14., in which it will 
be seen that the delicate exuvize of the real pupa are partially ex- 
serted out of the exuvie of the larva; the pupa, according to M. 
Wesmael (who has subsequently published a description of the larva 
of Subula marginata, dan. Soc. Ent. France, 1837, p. 91. app.), pos- 
sessing the power of pushing itself forward, previous to assuming the 
perfect state, by means of small ciliz directed backwards, and placed 
in rows at the extremity of each dorsal segment of the abdomen. 
The Rev. F. W. Hope has also given me specimens of Subula macu- 
lata and its preparatory ‘state; the latter found by him in a dry 
rotten tree in the New Forest, and from which he reared the imago. 
It is larger than that of X. varia, but does not otherwise differ 
from it. 
The transformations of Beris and Actina, the two other British 
genera of this family, have not been cbserved. 
* Although Latreille names this family Decatoma, “ten joints,’ he deseribes the 
antenne as 3-jointed, the last being diyided into “ huit anneaux.” 
