OF THE MOUTH PARTS OF CERTAIN INSECTS. 189 
of the galea a series of ridges which, under a high power, look extremely suggestive 
of the structures found in the labellz of Diptera, especially where, as for instance in 
Bombylius, the pseudotrachea are imperfectly developed. These ridges vary much in 
the species; but are particularly marked in a little Andrena near vicina, if not that 
species itself. Here we see (PI. III, Fig. 3) the entire inner face clothed with a thin 
membrane which is crossed by numerous closely set fine chitinous lines! I claim that 
this structure is the homologue of the pseudotracheal structure in the Diptera, and that 
in the latter order it is in the galea that the development occurs, as it does here in the 
Hymenoptera. The relative differences in size are not of importance. As to the 
particular use of this structure in Andrena I have no suggestion to make. 
In the Proceedings Ent. Soc. Washington, Vol. III, Mr. Ashmead figures on 
Pl. III, some very suggestive mouth structures of parasitic Hymenoptera, of which 
that of a Pteromalid is reproduced on Pl. III, Fig. 18. The central labium with its 
attached structures is much reduced in size, and the maxille, bearing the well-devel- 
oped palpi, are reduced to a single structure, the galea, resting upon what may be con- 
sidered the stipes. Now if we bring these two parts of the maxille a little more 
closely together, we have almost the exact structure seen in 2bio (PI. ILI, Fig. 11°). 
The basal ring, bearing the palpi, corresponds almost exactly to the basal ring of 
Pteromalus except for size, while except that the surmounting galea are two-jointed, 
the correspondence with the upper portion of the structure is equally marked. The 
labium in Brbio is much like that figured in Pl. III, Fig. 14, for Hermetia, and in PI. 
I, Fig. 12, for Huparyphus. 
I am making no very risky statement when I assert that the sclerite to which the 
maxillary palpi are attached must of necessity be maxillary; and further, it is equally 
safe to say that no maxillary sclerite can bear a labial appendage: and certainly not a 
labial palpus. It would be an absurdity, contrary to all the laws of a natural develop- 
ment, for a modified labial palpus to become attached to the sclerite bearing also the 
maxillary palpus; while if we consider it the two-jointed galea, its position is normal, 
requires no assumption of change or character, and does not differ in any essential 
points from the gale of the roach (Pl. III, Fig. 8). Yet these two joints in Bzbz0 
will, with a ridged membrane thrown over them, represent the labellate tip of the 
Muscid proboscis. That such a ridged membrane is well within the range of galear 
variability we found in the Andrena near vicina (PI. III, Fig. 3). 
The structure in Huparyphus bellus (Pl. I, Fig. 12) resembles Pteromalus yet 
more closely, in that a single ring only surmounts the segment bearing the palpus. In 
this instance the maxilla is reduced to exactly the same segments seen in the Hymen- 
opteron, and logic demands that we recognize them as the same. In this case, how- 
