238 CALIFORNIA ACADEMY OF SCIENCES. 



scholtzia, for instance — a very variable member of a variable 

 family which has lately suffered such great amplification — 

 we are already gravely asked to count the number of 

 stamens, and consider the fraction of a line in the width of 

 a torus, or the length of a filament. There are in this genus 

 of such wide distribution, probably fifty more species quite 

 as good as those recently described, and Platystemon prob- 

 ably contains nearly as many. 



That the flora of California is unusually variable is be- 

 ginning to be understood. Some of the causes of this 

 variability are well set forth in a preceding paper, and 

 probably amplifications of our well-known genera will 

 henceforth be looked upon with some doubt. The tendency 

 of natural history has usually been in the opposite direc- 

 tion — many species described, often from widely separated 

 localities, and from single or fragmentary specimens, 

 being found, in the course of time, to be too closely con- 

 nected by intermediate forms, the species are more or less 

 reduced. 



That the conclusions drawn by the writer from the fol- 

 lowing notes may in some cases be erroneous is quite 

 possible. That there may be no doubt as to the facts, sets 

 of the plants noticed will, as far as possible, be distributed 

 to the principal herbaria, especially to those most interested 

 in our flora. 



As will be seen by the following tribal arrangement, 1 

 have ventured to differ from Dr. Gray's last tentative revis- 

 ion by replacing Dendromecon in Hunnemannice,where, by its 

 aflSinities, it seems to belong; and from Bentham & Hooker 

 (though agreeing with Dr. Gray), by placing Romneya in 

 Eiipapavereoi. 



The general opinion (with which I fully agree) seems to 

 be that Hunnemannia cannot be kept up as a distinct genus, 

 but as it is known to me only by description, it is left for 

 future consideration. 



Tribe I. Platystemone^. — Leaves mainly opposite or 



