390 
writer is here used for the first time, and a brief explanation may 
be necessary. The endosperm is closely surrounded by the seed- 
coat and usually does not fill the entire carpel. In the group with 
lunate cross-section the endosperm is more or less grooved upon the 
inner face, thus producing the crescent-shaped appearance. In 
the other group this groove is greatly exaggerated. Not only is 
it very deep, but the sides are extended until they meet at the 
inner angle of the carpel, and thus the endosperm becomes a hollow 
sphere filled at first with the cellular seed-coat, but at maturity 
entirely empty except for the air which it contains, and which 
probably aids in floating the seed upon the water. A com- 
plete discussion of the fruit characters in the genus Galiwm will 
soon appear in a separate paper, and therefore nothing more — 
need be said here upon this subject. Curiously enough the type 
of inflorescence seems to supplement the fruit characters. In 
the writer’s opinion the number of corolla lobes offers a specific 
character. The plants are not “either” 3- or 4-merous, but the 
small three-parted obtuse lobed corolla represents one group of 
species as is shown by the supplementary characters of a smooth 
stem and a peculiar general appearance; while the four-parted 
large and acute lobed corolla just as accurately defines another 
group of species, and the character is here again supplemented 
by other characters. Valuable characters have been drawn from 
the arrangement of the flowers and the nature of the pedicels, 
while the leaves have not as a whole yielded good results. Very 
little dependence could be placed upon the number of leaves in a 
whorl. 
Linnaeus in 1753 described a species of Galium as G. trifidum, 
habitat Canada. It is necessary first of all to determine just what 
plant this name represents. Many of the earlier botanists were in 
doubt as to this point. Most of them, however, decided to in- 
clude under it all the obtuse leaved forms treated in this present 
paper. Hooker, however, having in hand perhaps mostly ma- 
terial from the States, and not having seen the true American — : 
type, considered the European form distinct, and to the American 
plant gave Michaux’s name G. C/aytoni. After this time the two 
forms were again united and have remained so until the present — 
time. It becomes evident, however, that the small bog form, the 
