182 AN ESSAY ON THE DEVELOPMENT 
there is no difference from Andrena or Polistes, which have been already studied. I 
have found no species which shows all the parts more fully developed than Xenoglossa 
prumosa (Pl. II, Fig. 15). Here all the parts are equally developed and all are func- 
tional; hence it makes a good starting point. The mentum is not shown in the figure 
except at the point to which the other parts are attached, and surmounting it cen- 
trally, we find the ligula; here a united, though extremely flexible organ. Lying cen- 
trally upon it, so as to close a groove, is the hypopharynx, in this case not easily separ- 
able from the ligula. Arising close to the central organ on each side are the para- 
gloss; almost as long as the glossa itself, flexible, unjointed, flattened and a little 
incurved at the margins so as to form, when closely applied to it, a partial shield for 
the ligula. Outside of all, situated at the outer margins of the mentum, are the palpi. 
These are four-jointed; but the basal joints are enormously elongated in proportion to 
the terminal two, and they are also flattened out, broadened and infolded, so that when 
at rest they cover and almost conceal the other labial parts, though not extending for- 
ward as far as they. In this insect the structures just described are almost entirely 
covered by the maxille, and a transverse section (PI. II, Fig. 15") is interesting and 
instructive. It represents the structure at about the middle of the combined maxillee 
and labium and illustrates the relative position of the parts. 
The tendency in the bees is toward a loss of the paraglossz, which shorten grad- 
ually until they disappear altogether, as represented in a species of Bombus figured in 
Pl. II, Fig. 15. Every intergrade is represented in any good series of bee mouth 
parts, and in their rudimentary condition, without function, they appear in Bombus sp., 
represented on Pl. III, Fig. 6. The palpi retain their unique development, and in 
the figure just cited are seen to be as long as the ligula itself, the basal two joints en- 
folding it almost completely, while the terminal joints are much reduced in size and 
set near the tip of the second joint, on the outer side. in other species these terminal 
joints are proportionately yet more reduced and are sometimes difficult to find. The 
essential point to be noted is that at their best development the paraglossz are not 
jointed and that they tend to complete obsolescence in the most highly specialized 
types. The palpi in Bombus require a little further examination: Reference to the 
figure last cited will show a short segment between the mentum and the first long 
joint, and this is membranous in texture. The mouth parts in Bombus are folded 
when at rest and the hinge is at the mentum; hence the necessity for some such pro- 
vision to enable the palpi to bend safely. 
Now let us assume that the ligula of this Bombus became rigid and chitinized, 
and that the edges of the palpi enfolding it became united to form a complete cylinder ; 
and then let us examine Hristalis tenax (Pl. ILI, Fig. 5) in the light of this assump- 
