HYMENOPTERA. — TENTHREDINIDA. HIS 
by Curtis as 4-jointed; but Hartig states that there is only a constric- 
tion, and not a distinct articulation, between the third and fourth joints. 
The cells of the wings are numerous; they have been described as having 
three marginal and two complete submarginal cells; and such is their 
appearance in X. pusilla; but an examination of the wing of X. Dahlii 
(fig. 71. 16.) will enable us to form a more correct notion of the ana- 
logies of these cells. The ovipositor of the female is exserted, and as 
long as the abdomen; it is inclosed between two compressed exter- 
nally pilose lamellz, and is membranous, with a strong horny central 
rib. From the descriptions of Dalman and Curtis, it would appear to 
consist only of a single piece; but I have ascertained that, in X. pusilla, 
it consists of two plates (jig. 71. 14. extremity of abdomen of @, with 
the saws extended), very slightly serrated, and thus agreeing in its com- 
position with the saws of the other Tenthredinide. The anterior tibia 
have two apical calcariz (fig. 71.19.) ; the posterior tibiae have two 
bristles on the external edge in the middle, although described by 
Curtis as having only one spine at the tips. Latreille, under this genus, 
states, “les larves vivent dans l’intérieur des végétaux, ou dans les 
vieux bois.” (Régne An. 2d ed. tom. v. p. 277.) Dalman considered 
this genus to belong to the Siricide (Uroceride) ; Curtis, although 
noticing thatit beautifully unites the Tenthredinide with the Uroceride, 
and that in its ample wings and large stigma it bears considerable re- 
semblance to Lyda, adds (in consequence of his not being acquainted 
with the true structure of the ovipositor), that “ it cannot be denied 
that the compressed oviduct brings it close to Xiphydria ;” he accord- 
ingly placed it in the family Xiphydriade Leach. Dahlbom, on the 
contrary, unites it with the family Tenthredines, as well as Hartig, 
who places it at the end of the family. The curious little genus 
Blasticotoma Filiceti Adwg appears to me to be allied to Xyela in the 
structure of the antennz, although the terminal joints are obsolete. 
The genus Cephus ZLaér. (Trachelus Jur., Astatus Klug) is as 
anomalous as Xyela, and equally serves to connect the Tenthredinidee 
and Uroceridz. The antenne are multiarticulate (21—28-jointed)*, 
generally thickened towards the tips ( fig. 71. 25.); the mandibles are 
robust and strongly and irregularly toothed; the labium is divided into 
3 elongated lobes ( fig. 71. 21.); the labial palpiare 4-jointed, the joints 
being irregular in size; the maxillz are bilobed, the outer lobes being 
* They have been described as only 21-jointed by Klug, Hartig, &c.; but in the 
subgenus Phylloecus Newm, they are 28-jointed, and filiform. 
