Ig SOUTHERN CALIFORNIA ACADEMY OF SCIENCES 
In the herbaria examined there was but one specimen of 
Acrolasia pectinata and none of A. nitens, Greene. I have in 
consequence omitted all reference to those two though I think 
that they may be valid species. I have supposed A. dispersa, 
Wats. and Al, integrifolia, Wats., to be the same species. 
The species of the genus Trachyphytum may be elassified 
according to their vegetative characters in the following order: 
A. aurea; gracilenta; affinis; viridescens; albicaulis; con- 
gesta; davidsoni; desertorum; dispersa; pinetorum; montana; 
but for study purposes the most accurate method is that 
adopted by Dr. Greene in the ‘‘Flora Franciseana.’’ The 
shape of the seeds is here the key to the grouping. The speci- 
mens examined would seem to indicate that the shape of the 
seeds in the different species is fairly constant, and this with 
the other specific characters help to definitely determine the 
rank of each. 
In one important particular I must differ from all authors. 
A. gracilenta is invariably described as having seeds in three 
rows. In this district the plant is fairly common. I have ex- 
amined large numbers of them in the fresh state and have 
invariably found only one row to exist. So far as I can deter- 
mine the herbaria specimens from elsewhere show but one row 
of seeds. 
There are but few exceptions to ‘‘seeds in a single row’’ in 
this genus; these will be specified later. 
The seeds of A.gracilenta may be taken as a type of which 
all the species showing the prismatic form of seed are more or 
less a variation. Those of gracilenta viewed from above are 
triangular prisms as broad as deep, grooved on the angles, the 
upper surface depressed below the edges. Perfectly sym- 
metrical seeds of this shape are only to be found towards the 
centre of the capsule, those near the ends being all more or less 
distorted. In integrifolia the seeds are deeper and more oblique 
at the ends. In some of the others the grooves widen out so 
as to form irregularly sided seeds, but the general tendency is 
to conform to this prismatic type even in those species of the 
albicaulis group where the seeds are irregular and very seldom 
show any grooves on the angles. The width of the groove 
varies in different species but this is too sight a feature to be 
a means of identification. Reference to the tubercles on the 
seeds I have omitted as all show it more or less under the micro- 
scope, so that it is difficult to say when they may be considered 
smooth. Pinetorum comes nearest to being smooth. 
TKGEDNA 
Seeds prismatical with grooved angles. 
Petals 2 lines long or less. 
Floral leaves concealing flowers............ A. micrantha 
‘ 
