238 ON THE GLOSSOPHAGIN &. 
is probably also of Vampyrine origin. The third division contains but a single genus, 
viz., Phyllonycteris. It is so near Brachyphylla that it would be easy to effect the 
transition and remove the genus to the alliance expressed by the term brachyphylline. 
It is akin, therefore, if not annectant, to the subfamily Stenodermine.* 
The material available for the study just completed was not large, and two genera, 
namely, Monophyllus and Glossonycteris, 1 have not seen. I have concluded from the 
published descriptions of G'lossonycteris that doubts can be frankly expressed concerning 
the validity of this genus. Perhaps not enough stress has been laid upon the effects 
of age in attempting to separate it from Anura. 
Reliable characters are found in the lower molars. The extension forward of the 
ridge (anterior commissure) between the protoconid and the paraconid is more marked 
than in any other group, and is in consonance with the compression of the crowns. The 
ridge is not spinose, and is scarcely raised. In Glossophaga the ridge is constantly as in 
the Vampyri, but in the other genera it is an extension forward from the protoconid. 
No trace of hypocone is seen in the upper molars. 
The row of glands lying to the outer side of the nostril is discernible in all genera 
except Phyllonycteris. Minute distinctions are found in the degree of development of 
these glands. They are best developed in the glossophagine group, and least so in the 
cheernycterine. In Phyllonycteris the ecto-nareal gland-row is occupied by a flattened 
fold of skin which becomes incorporated with the nose leaf.+ 
The proportions of the width of the third and fourth digital interspaces taken at the 
distal ends of the metacarpal bones when the wing is extended is found to be as valuable 
an aid in determining affinities as elsewhere in the order. In like manner the shapes of 
the terminal cartilages of the fourth and fifth digits, the arrangements of muscles and 
nerve markings of the wing membrane are noted as furnishing excellent characters. 
The following scheme of interdigital diameters is given : 
Second Third Fourth Second Third Fourth 
Interspace. Tnterspace. Interspace. Interspace. Interspace. Interspace. 
Glossophaga soricina .....- Q 12 17 = Lonchoglossa.. cog 16 23 
Glossophaga truei........-.. 2 11 15 JAGR acconscqcon0nc0ss9008002 3 15 30 
Eeptony Ctr is ...-.eseeeeeee 3 15 25 Phyllonycterts ....+++...... 3 13 25 
Cher ony cteris........-000+ 2 11 20 
Enough can be gleaned in the way of inductions from the shapes of the anterior 
* In a paper by myself, entitled ‘‘On Ametrida minor” (Proc. Bost. N. Hist. Soc., 1892), I used inadvertently the 
term Stenodermatide for this subfamily. 
+ The genera of the remote megaderminine genera are in like manner distinguished by characters in rows of glands 
as contrasted to folds of skin, though the structures are here not ectonareal, but infranareal. In Megaderma the glands 
are distinct, while in Lyroderma and Lavia they ave supplanted by a skin-fold which becomes an integral part of the 
nose leaf, 
