72 
REVIEWS. 
Subula. The characteristic of this family gives no information as to the 
extent in which Mr. Walker takes it. Something should have been said 
of the veining of the wings, in which the relation of the family to the 
Leptidce , Xylotomce , and Tdbanidce is so clearly expressed. It should* 
also, have been remarked that it is distinguished from the Leptidce chiefly 
by the flagellum of the antennae, composed of a greater number of closely 
connected joints; from the Tdbanidce , by the different structure of the 
mouth; from the Xylotomce or Thereuidce , by the development of the 
empodium in the form of a pulvillus ; while the spurred tibiae and the 
ambient costal vein separate it from the Stratiomydce 1 to which it makes a 
nearer approach through the genera Arthropeas and Coenomyia. On the 
other hand, the characters taken from the structure of the antennae, which 
are applicable only to the two British genera Xylophagus and Subula , 
might well have been left out. This is one of many instances which 
clearly show the inconvenience of the method followed by Mr. Walker in 
the determination of the family characters. Supposing—what is by no 
means impossible—that Coenomyia ferruginea should be hereafter discovered 
in Britain, this will require a totally different character of the family ; and 
that again will have to be altered in case Pachystomus syrphoides also 
should occur. Where will this end, if the scaffolding on which all future 
discoveries are to be arranged has so little stability that the occurrence of 
one new genus, and even in some cases the accession of a single species, 
endangers its utility and makes a multiplicity of amendments necessary ! 
What would be said of a flora in which the classes and orders were 
characterized after such a fashion ? 
3. TABANiDiE.—The wings cannot properly be termed “ deflexse,” 
Hcematopota being the only genus in which they are actually so. The 
antennas are said to be six, eight, or ten-jointed: in all the genera here 
given they are either six or seven-jointed ; in Chrysops , indeed, the third 
joint is annulated, and seems to be composed of several joints, but these 
are completely connate. The palpi should have been given as two-jointed 
in this family. An <£ obconical abdomen” (conical the author means) is 
attributed to the males—-a character not applicable to the genus Chrysops. 
4. Acroceridjl —The size and form of the alulas, which completely con¬ 
ceal the halteres, in a manner especially characteristic of the family, should 
have been noticed ; also that the eyes are contiguous in both sexes. The 
antennas are described as three-jointed, which is not true of either of the 
British genera; the seta in Acrocera , and the style in Henops , being a 
simple process of the second joint. The number of the ocelli, given a® 
