PIPER—BERBERIS AQUIFOLIUM AND BERBERIS REPENS. 447 
In 1916 Dr. Albert Mann made a careful microscopic examination 
of the leaves of three variants of B. aquifoliwm from the type local- 
ity. The dull-leafed form was then supposed to be B. repens and 
the plant is thus named in his report, which is as follows: 
“A microscopical examination of the leaves of Berberis repens and B. aqui- 
folium gives no satisfactory distinctions on which one could base a claim for 
difference of species. Transverse sections of these leaves, made at the same 
place in the lamina, show that aquifolium has the same number of rows of 
palisade cells as repens, but the cells are smaller and consequently there are 
more in number in a given area; the spongy parenchym is not so loose, and 
has thicker walls; the cuticle, especially of the upper epidermis, is thicker. 
This difference could readily be accounted for by a difference of habitat, as 
these characteristics are strongly influenced by light, and especially by a 
minimum quantity of moisture. 
“A comparison of the epidermal surfaces torn from the leaf shows more 
marked distinctions, but hardly justifying a specific separation. The upper 
epiderm of aquifolium consists of smaller cells than that of repens, far less 
serpentine in outline, quite thick, and abundantly cut across by canaliculi; 
those of repens quite sinuous, thinner, and only slightly, if at all, perforated. 
The under epiderms contrast even more strongly. The stomata of aquifolium 
are more abundant and with larger guard-cells, with occasionally lenticels on 
the surface, the other cells of the epiderm being considerably smaller—from 
two to four times—than those of the stomata. The cells of repens are from 
one to one-half the size of the stomata, and the walls are, as in the upper 
epiderm, very much thinner. 
“TI. do not see how any specific distinction can be drawn from these con- 
trasts, as I am not at all sure that if the habitats of the two plants were 
exchanged, the differences noted by me would not be reversed and almost 
as marked as above recorded. 
“There is a striking difference in the coloration of the leaves, which results 
in the stomata of aquifolium standing out from the rest of the epiderm as 
deep brown chlorophyll-containing cells; but this is doubtless due in part to 
the greater age of the specimen of this species.” 
Since the receipt of Dr. Stapf’s letter, numerous specimens have 
been examined to test the value of the character that he points out, 
and it seems that in all critical cases it furnishes a definite basis of 
determination. The lower epidermal cells in both species project on 
their free surfaces as low papillae. In B. repens these papillae, as 
viewed vertically, are small, circular, prominent, and distinctly 
separated; in B. aquifolium they are lower, larger, and contiguous, 
thus assuming almost exactly the cell outline. A reexamination of 
Pursh’s type specimen on the basis of this distinction alone, places 
it definitely with the shiny-leafed species. 
On the basis of the specimens examined, Berberis aquifolium 
ranges from Vancouver Island and southern British Columbia south- 
ward to the Callipooia Mountains of Oregon and eastward to western 
Idaho. The only other species in which the lower epidermis has a 
similar structure is Berberis pinnata Lag., of the coast region of 
