218 BOTANICAL GAZETTE [SEPTEMBER 
as in Frenelopsis (30), or they may make a rim around the opening 
itself, as in species of Taxus, Juniperus, Thuja, Libocedrus, etc. 
In other cases, all the epidermal cells, at least in the vicinity of 
the stomata, may have peculiar knoblike projections of cuticle. 
These last seem to be diagnostic in their constancy; they are 
present in Taxus baccata, including vars. erecta, fastigiata, and 
variegata. 
As a last feature, we may refer to the shape and extent of the 
lignified lamellae of the guard cells. These have been found to be 
absolutely constant in the case of the cycads, and the investigations 
of the writer on the conifers indicate a considerable uniformity 
not only within the species, but even within the genus or family. 
Thus in all the members of the-Araucarineae examined, the ventral 
thickenings are relatively larger and overlap the dorsal to a greater 
extent than in any other family. Unfortunately, however, the 
employment of this feature in examining fossil conifers is usually 
rendered impossible through imperfect preservation, although 1 it 
has been found to be of considerable value for the cycads. 
From this discussion of the significant points in structure of 
the epidermis, two conclusions may be drawn. The first is that 
the number and general distribution of stomata (upper or lower 
surface of leaf) is variable within the species, but that the character 
of the walls of the epidermal cells (uniformity of thickness, cuticular 
projections, straight or sinuous), arrangement of stomata (whether 
or not in rows, angle of long axis with reference to leaf margin), and 
extent of lignified lamellae of guard cells, are remarkably constant. 
The second relates to the phylogenetic value of these features and 
seems to indicate that it is small. For example, all leaves of the 
Taxites type, no matter to what family they belong, have long 
rows of stomata on each side of the midrib of the under surface 
of the leaf all with their long axes parallel to the margin. Similarly, 
leaves of the Thuyites type, as exemplified by Dacrydium, Arthro- 
taxis, Tetraclinus, Thuya, Juniperus, Cupressus, Libocedrus, etc., 
have the stomata scattered indiscriminately, usually avoiding the 
midrib, but with no sign of rows or constant angle. It seems 
obvious, accordingly, that the investigation of cuticles is of impor- 
tance to the systematic botanist engaged in accurate specific diag- 
