POST-EMBRYONIC DEVELOPMENT OF CULEX. 175 
of unjointed dorsal appendages of the thorax, forming the 
siphons of the pupa and the wings and halteres of the 
imago, three pairs of ventral jointed appendages of the 
thorax, the legs of the imago, one dorsal pair on the eighth 
segment of the abdomen, and two ventral abdominal pairs 
in the composite mass called for convenience ‘‘ ninth seg- 
ment’ of the abdomen, and probably formed of three 
segments. These two last are the gonapophyses of the 
imago—appendages of the supposed eleventh segment I 
have failed to make out in the larva. All these eight pairs 
arise alike as outfoldings of the epidermis (‘‘ hypodermis’’) 
into which mesoblastic tissues extend. They all lie upon 
the surrounding epidermis under the cuticle, and are vari- 
ously folded according to the form of the appendage itself, 
and in the case of the six thoracic pairs they are lodged in 
depressions of the body-wall. 
Of the dorsal thoracic appendages all three are at first 
plate-like, but the first pair soon roll up to form the tubular 
siphons. he rudiments of the halteres and of the wings 
differ chiefly in point of size, the wings being much the 
larger. The legs are more or less cylindrical from the first 
but much crumpled as shown in Weismann’s figures of 
Sarcophaga (1). The dorsal appendages of the eighth 
abdominal segment, the future fins of the pupa, are 
plate-like bodies lying immediately beneath the cuticle 
of the larval siphon. The rudiments of the gonapophyses 
are rod-lke bodies lying under the cuticle at the sides of 
the “ninth segment.” 
Of the mouth appendages it is more difficult to speak 
with certainty, for the cuticle of the larval mouth-parts 
is so hard that it is almost impossible to cut sections of 
this region. ,Sutfice it to say that when the pupa has 
emerged from the larval cuticle, the head-appendages are 
all very much longer than the larval head they have come 
