154 
lir nieces between which the unusually developed thirteenth seg¬ 
ment anpears These lateral portions are fringed with spines on 
the posterior two-thirds of their inner margin, and the segment is 
lonXdinallv divided beneath. These two _ segments together are 
about one and a half times the length of the preceding. Their 
dorsal portions are likewise elongate, as compared with the er 
“s being scarcely, if at all, shorter than the next segment m 
advance. 
In the form of the mouth parts, this root-worm is likewise distin¬ 
guished from at least one of the others, the tips of the mandibles 
margins are uniformly cuived. 
It is in these latter structures only that any character can 
be found by which the larva of Scelodonta may be separated 
from that of Paria. In every other minutest particulai of form and 
structure even to he number, distribution and arrangement of the 
hails “non the surface, these two larvse agree precisely. But 
while the tips of the mandibles of Scelodonta aie either entire, 
or if notched, are so lobed that the central division is much the 
longest those of Paria (Plate IX, Fig. 5, B,) are obtuse at the 
extremity and distinctly emarginate, being sometimes deeply bifid. 
It must be confessed, however, that none but an expert will be 
able to distinguish between these two species in the larval stage, 
and it is not fmpossible that this trivial character will be found too 
variable to be depended on, except where a considerable number of, 
specimens are studied together. _ 
The pup® may be separated by differences m the armature o 
the posterior segments of the abdomen especiaily the las^lhi* 
segment in the pupa of Golaspis, (Plate IX, ii§- -> »/ | 
in g twJ staple hool, the poinfs of which curve towards each oth^ 
while in both Paria (Fig. 5, A) and Scelodonta, (Fig. 1, A) these 
hooks are larger and longer, and curve dorsally instead of lnwaids. 
In front of these hooks are two pairs of lateral spines, one attac e 
to the ultimate segment and the oilier to the peuultin: > 
mer of which in Golaspis extend directly outward, while in the other 
larvse they project backward instead. 
There are likewise slight differences in the spines upon the lmee# 
of the pupa. In Colaspis a strong curved hook arises from t ft 
anterior inferior angle of the tibio- emoral articulation of all the 
legs, that of the anterior pair being, however, very small. In b «“° 
donta and Paria, on the other hand, we have a stout spine upon 
the posterior legs, but the corresponding angles of the two anterior 
pairs are unarmed. I 
The most careful comparison of the pupae of the two m01 ,® c 
related species, has enabled me to detect only a single dfferencj 
between them. In Scelodonta (Fig 1, A), the terminal ! abd(innnai 
hooks have each a strong, erect tooth or spine attached t° PP 
side of the base while in Paria (Fig. 5, A) this spine is wanting. 
k 
! 
