( Hi ) 
their anthers fo me what imperfect. As the 
fructification advances towards maturity, the 
filaments continue to enlarge about the mid- 
dle, while the top is drawn out to a kind of 
beak, in which {late they- might be m,iftaken 
for the piftil. On cutting them through they* 
appear hollow, and to contain fome of the 
fame corpufcles, which may be feen on the 
outfide of many of them, fo that it would 
be impoffible to know them to have been 
originally filaments. This mows yo\i the ad- 
vantage of examining flowers in their dif- 
ferent ftates of maturity, and before the full 
expanfion of their corols. The fempervivum 
is nearly allied to the fe'dum, but differs in 
having more than- five petals ; it is alfo liable 
to increafe in it's number of piftils, when it 
grows luxuriant. 
We are obliged to Mr. Curtis for an accu- 
rate knowledge of the difficult and curious 
genus Euphorbia, which is the botanic name 
of the churn-ftafF. He juftly remarks^ that the 
Linnean characters of this family will not, in 
any of the Britifh fpecies, even guide us to it's 
clafs. The flamens are very minute ; there are 
feldom more than two or three that appear 
above the calyx, the reft are concealed within 
