*73 
of being bent outwards and backwards. If we now flex the fingers 
this gives approximately the position that the ventral half of the plate 
(the teeth) takes when the proboscis is everted. 
The two bilateral inner walls are at first approximately parallel. 
If we now trace them backwards, we find that at first the inner 
walls approach one another ventrally so that they now enclose a 
somewhat V-shaped space (fig. 22). The dorsal half of the 
plate (£) grows downwards at the expense of the ventral, so that 
in fig. 21 we have only a few remnants of the ventral half left 
(and here the two sides are united ventrally), and eventually these 
disappear, so that we have on either side what was the dorsal half 
now occupying the whole inner side (fig. 22, k). These two sides now 
become fused ventrally by a median structure (figs. 23. 24), the 
axial apophysis (if), which serves to bind them together. 
Dorsally the inner sides now become replaced by new structures 
appearing, viz., by two lateral chitinous rod-like sclerites, seen in cross- 
section in figs. 21-24, r, and medially a structure (r 1 ) appears which 
will form eventually the floor of the labial gutter when these two rods 
and the median piece fuse to form a single trough. 
Labial gutter (r, r 1 in all the figs.). I he commencement of the 
labium can be seen in a cleared proboscis as a transverse line stretching 
across (above) the middle of the axial apophysis. In sections it appears 
at about the level of the tips of the fork as a slightly hollowed-out free 
chitinous piece. About the same time appear, dorsal to it, two 
chitinous rods (r) seen in cross section in figs. 21-28. These rods 
are hooked at their edges, and to these edges are attached ligamentous 
tissues continuous with the chitin externally. At first there is only a 
fibrous junction with the ventral piece, which soon expands laterally 
and dorsally, and it is not till some way down that the fusion takes 
place and the canal assumes its complete shape. In the region of the 
base of the fork the canal has developed a ventral projection (fig. 28), 
and this goes on developing until the appearance of a row -lock is 
obtained, as seen in fig. 29, which again disappears in the bulb. 
The axial apophysis expands posteriorly into a rounded mass seen 
in fig. 25, and by this level the original lateral internal walls of the 
labella have completely disappeared, the hind portion of the apophysis 
(<7) now' lying free, imbedded in soft tissues. 
