De a en ee eee eo \ i ee ee ee 
i i RL Se Fa eh eee es a 
AGAVES OF THE UNITED STATES. 65 
stigmas, which slightly open or broadly expand at maturity ; 
fruit a dry, erect, globose to cylindrical capsule, loculicid- 
ally three-celled, having two rows of numerous, thin, black 
seeds in each cell, and generally opening in upper part 
only; embryo filiform. 
The inferior ovary* gives the strongest reason for class- 
ing this genus with the Amaryllidaceae. Its capsules, 
numerous discoid seeds, and elongated cotyledons show 
affinities with Liliaceae. Within the Amaryllidaceae, it is 
closely allied to Furcraeae. 
The best basis for a classification of the Agaves lies in 
the fundamental differences in the forms of navoreseohiGn: 
accompanied, as they are, by group differences in the struct- 
ure and forms of the leaves. The sections recognized by 
Dr. Engelmann, — Singuliflorae, Geminiflorae, Panicu- 
latae, may be very technically described as having flowers 
usually subspicate and solitary; flowers usually subspicate, 
in pairs; and flowers paniculate. As these are subgeneric 
* Leichtlinia protuberans Ross, A. protuberans Engelm., has been 
placed between the genera Polianthes and Agave, on account of its coni- 
cal ovary protruding into the perianth. 
On July 29th, I collected a monstrous inflorescence of A. applanata 
Parryi, in the mountains above Pleasant Valley, a few miles from Fort 
Bayard. The top of the scape had been broken by some accident, and 
the plant had made an effort to produce flowers on a low branch of the 
inflorescence. ‘These flowers were in a thick mass close to the main 
axis. All were imperfect or distorted. Some were grown together. 
The segments in nearly all cases were greatly broadened and frequently 
thickened. The filaments also were broad and in some cases showed a 
distinct reversion to the petaloid character. In some flowers it was 
difficult to tell whether a certain organ represented a segment or a fila- 
ment, but in the larger and better developed flowers, there was usually 
an equal number of each, and this number varied from six to five, four, 
three, and even two. In one large flower the style was irregularly four- 
lobed, and the stigma, three-lobed, one lobe being much larger than the 
other two. The ovary was usually represented by a short thick mass of 
tissue with little or no differentiation. Mr. Webber writes me of finding 
a monstrous Agave flower upon a plant of what I suppose was A. rigida 
sisalana. This flower had stamens and pistils perfectly developed, but 
was without any ovary differentiation, and was found growing from a 
cluster of leaves of the bulb. 
