36 Evans: HEpATICAE OF PUERTO RICO 
short branch, bracts usually in three to ten pairs, closely imbricated, 1 
strongly inflated, shortly bifid, the lobe acute or obtuse, the lobule ~ 
subacute, keel strongly arched, narrowly but distinctly winged in — 
the upper part, the wing slightly denticulate from projecting cells; 
bracteoles present at base of spike, small, bifid with erect-spread- — 
ing lobes and obtuse sinus ; antheridia in pairs, (//. 5). ; 
On living leaves. El Yunque, Evans (21 p.p., 23p.f.). Orig- 
inally collected by Spruce on Chimborazo. i 
The type-specimens of Lefeunea infundibulata in the Spruce — 
herbarium are exceedingly fragmentary and are badly mixed with | 
L. campanulata Spruce, a closely allied species. They are at the i 
same time so weather-worn that it is difficult to gain from them 
a clear idea of the species. Fortunately Spruce’s description is i 
definite and is apparently drawn from better specimens. The — 
plants from Puerto Rico are in good condition and agree closely 
with the original description and also, so far as one can determine, 
with the battered type. A single discrepancy with regard to the 
keels of the perianth may be noted. According to Spruce these — 
are ‘‘ dilated into broad, acute, almost horizontal subula’’ and it is _ 
implied that they are smooth. In the writer’s specimens the keels q 
extend outward as slightly spinose horns. An examination of — 
the type shows, however, that even here the keels are not invari- ‘ 
ably smooth but are sometimes more or less toothed, so that the 4 7 
species is apparently subject to considerable variation in this respect. 
D. infundibulata is very closely related to the African Lejeunea 
capulata Tay|.* but differs in its exocellate leaves. According to 3 
Spruce + L. capulata occurs in the Amazon region, but the speci- a 
mens in the Spruce herbarium referred to this species, although _ 
very fragmentary, are evidently distinct from the type-material of : 
L. capulata in the Taylor herbarium. They seem to be much 
closer to D. bidens or to L. lichenicola, but it would be unsafe to 
refer them definitely to either. At all events the evidence is very 
inadequate that L. capulata is a member of the American flora. 
Other species more or less closely allied to D. infundibulata | 
are Lejeunea campanulata, already mentioned, and the Puerto 
Rico species, D. diocellata and D. dissitifolia. L. campanulata— 
* Lond. Jour. Bot. 5: 394. 1846. 
t Hep. Amaz. et And. 1go. 1884. 
