102 BOTANICAL GAZETTE. [AUGUST 
majority of the small rays are less than 20 cells in longitudinal 
extent, and are one or two cells wide, often two cells wide in the 
middle and one cell wide at both ends (fig. 12). Radially, the rays 
extend from the pith to the cortex, few if any new rays being 
formed as the trunk grows. Most of the cells contain large starch 
grains, but some have crystals of calcium oxalate, which is also 
abundant jn the phloem, cortex, and pith. 
The large rays also extend from the pith to the cortex. Longi- 
tudinally, they measure 4 to 8 mm., and from a width of about 
1 mm. in the middle they taper to a single cell above and below. 
In each large ray is a leaf trace, with its phloem more or less dis- 
organized. The xylem of this bundle is usually uppermost, but 
the orientation is various until the bundle reaches the cortex, 
where it becomes a part of the characteristic girdle. From both the 
pointed ends, tracheids extend into the ray, often making nearly half 
of the ray look like a group of small rays (fig. 12). Most of these 
tracheids which extend into the rays are scalariform, but some are 
slightly pitted, and some are the cells of the wood with transverse 
walls already described. Every large ray has at least one mucilage 
duct, and surrounding it at a distance of a few cells, the calcium 
oxalate crystals are particularly abundant. 
While the large ray, especially in tangential section, resembles 
the broad ray of Quercus, as described by EAmeEs (9), its mode of 
formation is different, the broad ray of Quercus originating by the 
fusion of small rays, while in Dioon the broad ray owes its origin 
to the leaf trace which it contains. The tissues simply grow around 
the leaf trace, and the compound appearance of the ray, shown at 
the upper end of the photomicrograph (fig. 12), is a secondary, not 
a primary feature. 
COMPARATIVE HISTOLOGY.—The trunks of Dioon spinulosum, 
D. edule, Ceratozamia mexicana, and Zamia floridana have some 
histological characters in common, and it is probable that all the 
Cycadales have enough histological peculiarities to identify the 
order by the structure of the trunk. The four species mentioned 
above are easily distinguished from each other by such histological 
characters, but it is very doubtful whether nearly related species 
of a large genus like Zamia could be so distinguished from each 
