HYMENOPTERA — APIDA. 269 
gathering bees (Anthidium); 5. Solitary parasitic bees (Nomada, 
Pasites, Melecta, &c.); 6. Solitary woolly-legged bees (Eucera, An- 
thophora, Centris, &c.); 7. Bees temporarily social (Euglossa, 
Bombus); 8. Bees permanently social (Apis, Melipona, and Trigona). 
In this work Latreille considers the parasite bees more nearly allied 
to the woolly-legged ones (by means of Melecta and Anthophora) 
than to the Andreniform bees (by means of Nomada and Panurgus) 
as arranged by Mr. Kirby. In effect both these affinities seem equally 
true, the difficulty as to the arrangement of these parasitic genera in 
accordance with their real allies, arising from the parasitic nature of 
the former, and the solution of the question whether their arrange- 
ment ought not to be entirely independent of such consideration. In 
his last general work, the second edition of the Régne Animal, La- 
treille has advantageously reduced the number of primary groups, 
and divided the family into five sections: —1. Andrenoides ( Andrena- 
like bees, including the first group in his Genera, with the addition of 
Xylocopa); 2. Dasygastres (hairy-bellied bees, including the re- 
mainder of the second, and the third and fourth groups); 3. Cu- 
culinae (cuckoo bees, his fifth group); 4. Scopulipedes (brush-legged 
bees, his sixth group); 5. Sociales (social bees, containing the two 
remaining groups of his ‘Genera’ ). 
This arrangement [ shall adopt with the follasetne slight alterations : 
the subfamilies Dasygastres and Cuculine, being established upon 
characters arising from their polliniferous or parasitic economy, will 
(for the reasons already given) require modification, as indeed La- 
treille himself admits. (Ftegne Animal, tom. v. p. 347.) For the former 
I therefore propose the name of Longilabres, long-lipped; and the 
latter, which, however, I would only provisionally retain, I denominate 
Denudate, naked bees; restricting the limits of these groups by the 
characters which have supplied their altered names. Moreover, the 
genus Xylocopa appears to me to be much more nearly allied to An- 
thophora* and Centris, than it is either to Ceratina or Panurgus (as 
indicated by Latreille), or to Bombus, as insisted upon by Mr. Kirby. 
If we therefore adopt the following position of these subfamilies, we 
shall perhaps obtain an arrangement founded upon a greater number 
* Since the above was written, I have become acquainted with, and described a 
remarkable osculant genus, proving this relation (Mesotrichia torrida Westw., in 
Trans. Ent. Soc. vol. ii. pl. 9.). 
