190 Annals Entomological Society of America  [Vol. Son: 
XIII), often reaching nearly to the base of the tergite on the middle 
line. In Pipiza (Figs. 17, 23, 26, Plate XI) and Spherophoria (Figs. 1, 
60, 63, 68, Plate XV) it is remarkably deep and in some species of 
Spherophoria it has completely closed caudad of the cerci so that the 
latter are embedded in tergite ten. In the Milesine and Eristaline 
it is relatively slight, but has the margin often somewhat elevated, 
rim-like (Figs. 90, 95, Plate XVI). In some genera (e. g., Chrysogaster, 
Figs. 36, 42, Plate X), it is practically undeveloped. 
Tue Cercr (acrocerci, Berlese; appendage iv, Newell; lamellz, 
Lundbeck; epiproct, Crampton; forceps superiores, Wesche) are the 
simplest and least variable of the appendages. They are typically 
one-segmented, sessile, semi-oval, compressed and lie closely parallel at 
either side of the anal opening. They are covered with minute micro- 
trichia and also scattered, longer hairs, usually arising from rather 
prominent basal rings and doubtless tactile or otherwise sensory in 
function. Since they are so closely incorporated into the structure of 
the tenth tergite, it has seemed best to describe their directions and 
margins like those of this sclerite. Their length, therefore, is the dis- 
tance parallel to the mid-longitudinal line of the tenth tergite and their 
width the transverse distance perpendicular to this. In cases where 
they become elevated and protruding, their heighth is the maximum 
distance more or less perpendicular to the plane of the tenth tergite. 
A part of the species of the genus Chrysogaster shows a peculiarity of 
the cerci not noted elsewhere in the entire family for they are sometimes 
(for example, C. pulchella), distinctly two-segmented. (Note Fig. 36, 
Plate X). In the other species they are one-segmented; but in several 
of them (for example, C. pictipennis), very distinctly emarginated about 
the middle. So that it is possible, in this genus alone, to arrange a 
series in which the cerci show a very gradual transition from the one- 
segmented, not emarginated condition, through various degrees of 
emargination to the distinctly two-segmented condition. If we con- 
sider this a stage in the specialization, by fusion, from an earlier multi- 
segmented condition, it places this genus as the most primitive of the 
family in this respect. But if we assume that the cerci in this group 
have, after an ancestral reduction to a single segment, again become 
divided, due to the emarginating tendency (noted also in other genera) 
it falls in line with the otherwise much specialized condition of the 
genus. 
In correlation with the slighter development of the cercal emargina- 
tion in some genera is the greater elevation of the cerci, either on their 
more basal end or on a slightly chitinized, bare, membranous petiole. 
In Eristalis, Tropidia, Syritta, Ceriodes and especially Xylota, they 
become very prominent relatively to the styles (see Figs. 91, 95, Plate 
XVI; 106, 110, Plate XVII; and 129, 131 and 132, Plate XIX), and the 
hairs, especially toward the apex, are larger and bristle-like. In Micro- 
don they stand erect on a slender base, expanded, somewhat bilobed at 
apex. They are unusually poorly developed in Brachyopa notata. 
