117 
missing sepal, it appears on the opposite side of the flower, there 
replacing what might have been a fifth stamen. We have then, 
so far, four sepals, one petal and four stamens, making collectively 
nine parts of the flower; and the symmetry is completed, up to 
the requisite number of fifteen parts, by the pistils being, as they 
almost always are, just six, of which five are inserted in a whorl 
around the margin of the minute receptacle, and the sixth sits 
centrally and apart from the rest on the summit of it. Such is < 
the ordinary number and arrangement of the parts of the flower 
in this most interesting little plant. I have observed, however, 
not a few instances of deviation from the rule; but these have 
been what may be called, if the paradoxical phrase may be 
allowed, very regular deviations from the rule, not affecting the 
numerical total of fifteen parts to each flower. There may be 
five pistils and two petals, the stamens and sepals each remaining 
four; or five pistils, five stamens, one petal and four sepals. 
When the petals are two, one of them takes the place of the 
missing sepal and is, in some respects, intermediate between a 
sepal and a petal, being uncommonly large for a petal of the 
species, and destitute of any trace of the nectariferous pit. In 
this case, too, the second petal is reduced in size, and appears like 
a kind of half-way affair between petal,and stamen. The normal 
solitary petal seen under a good magnifier consists of rather dis- 
tinct lamina and claw, with a comparatively large nectariferous 
pit (without scale) located at base of the lamina, which last named __ 
part about equals the claw in length. The plant is strictly 
annual, very slender, weak and often reclining, growing in shady 
and rather wet places, often in company with Montia fontana. 
The typical R. hebecarpus inhabits the Sierra Nevada, rather 
beyond my reach. It is quite different in appearance, and, 
although much cannot be learned about its flowers from the dried 
specimens, its akenes are many and quite indefinite, forming a 
close round head. For the variety here described I dare not yet 
propose that specific name and rank which it will some day com- 
mand, if the other fail to exhibit more or less of its really won- 
derful characteristics. I wait for some possible coming oppor- 
tunity of comparing the two in living specimens. . 
From this, perhaps, smallest of all Ranunculi, I pass to give an 
