fo * BOTANICAL GAZETTE, 57 
_herd’s purse, and it may help to explain the absence of 
_ 0dspores in this host. It is not unreasonable to suppose that 
a form of fruiting may be omitted when it is not essential for 
the continuation of the species, as has been shown to be true 
in species of Uredinew. There was an abundance of the 
Cystopus on Capsella early this spring, because the late rains 
of last year permitted the seedlings to make a good growth, 
and become thoroughly infested before the season closed. 
This spring these same plants grew and produced a large 
crop of spores, which probably would have rapidly spread to 
other plants had not the dry weather prevented. In June 
there was very little of the mildew. It was fairly abundant 
upon Lepidium Virginicum in early spring, but soon disap- 
peared. Two years ago, in a fairly moist year, the inflores- 
cences (flowers, seed-vessels, etc.) of the garden radish were 
‘very generally attacked, and often distorted beyond recog- 
ition. This season none of this mildew was found in simi- 
lar situations. Late in October it was collected in quantity 
upon young plants of Sisymbrium officinale, which had de- 
veloped in a moist shady place, after the rains of early Sep- 
tember. . 
Cystopus cubicus Lév. is the least common species of the 
senus, and for the last two years has been comparatively 
Tare. At distant intervals it was found upon Ambrosia arte- 
Mmisiefolia, 
Cystopus Blité Lévy. occurs upon an increasing list of hosts. 
Up to the present year it was known in the United States on 
Amarantus hybridus, A. retroflexus, A. blitoides and Acnida 
Canmnabina. ‘The additional hosts for 1887 are Amarantus 
albus and Montelia tamariscina. Upon both these new hosts 
the mildew was far from rare. The Montelia grows only in 
low moist ground, and what effect a wet season may have 
upon the abundance of its parasite remains to be determined. 
£ € host most commonly infested is Amarantus blitoides. 
This low-spreading weed grows in dry places, even upon 
Stavelly paths and roadways. It is, however, a thick, rather 
=4Cculent-leaved plant, and, like the garden purslane, is 
| Itself full of moisture, even though the surroundings are arid. 
Nevertheless, it was observed that the greatest development of 
mildew a peared upon plants which weresmost favored 
with moisture and shade. ree. 
ystopus Portulace Lév. is the last species of our list, and 
One of no little importance in its bearing upon the question 
Mhand. At first sight, it seems to give evidence contrary — 
