194 PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
racic limbs, respiratory trumpets, wings, halteres, and gonapophyses 
are developed from shallow invaginate buds. The labrum is formed 
from an evaginate bud. The thoracic and antennal buds are evagi- 
nated immediately prior to the pupation moult (pi. 16, fig. 48), the 
labial and maxillary palp buds after the larval cuticle is partly 
removed. The buds for the gonapophyses are situated on the ven¬ 
tral face of the ninth joint of the abdomen. The patches of altered 
hypodermis which represent in a rudimentary form buds for the tail 
fans of the pupa are formed low down on the sides of the eighth 
segment of the abdomen. In the species of mosquito studied by 
Hurst these buds are described as “plate-like bodies lying imme¬ 
diately beneath the cuticle of the larval siphon.” They lie high 
up on the sides of the segment in the larva of Anopheles, and in 
Corethra are almost dorsal in position. These differences are prob¬ 
ably correlated with the varied size and shape of the eighth abdomi¬ 
nal segment in the three genera : wide and unincumbered in Corethra, 
narrower in Anopheles with a small dorsal respiratory siphon, and 
very narrow in Culex with a huge dorsal siphon. The tail fan buds 
are the last to appear in the Culex larva. The order for the devel¬ 
opment of the imaginal buds is: antennal and thoracic; labial and 
maxillary palp ; labral; gonapophysial; mandibular and maxillary; 
tail fan. 
The antennal buds are of especial interest. They are deeply 
invaginated into the head cavity, and when completed the base of 
the new antenna at the bottom of the invagination pocket lies behind 
the eye spot (pi. 15, fig. 29, ant bcl). Miall and Hammond (’ 92 ) 
compare the type of antennal and eye development presented by 
Culex, where the eyes are maintained independently of the antennal 
invaginations, with that found in Chironomus or Simulium where 
invaginations give rise to both antenna and eye, and finally with the 
extensive invaginations of Musca from which eye, antenna, and 
the whole head are formed: the so called “brain appendages.” 
On a basis of this comparison he places Culex at the bottom of a 
series, Simulium and Chironomus higher, and Musca at the summit. 
This arrangement is perhaps justified as a statement of progressive 
elaboration of a tendency to form eye and head from an invagina¬ 
tion that is connected with the antennal bud. But it does not 
sufficiently accentuate the mechanical aspects of the phenomenon. 
Kellogg (: 02) appears to me to express the case better when he 
