470 Mr. A. W. Waters on 
material advance has been made, for, besides other characters, 
we now recognize that the ovicells are most useful by their 
shape and that of the oveciostome ; another useful character 
: the longitudinal distance from zocecium to zocecium, as 1S 
also the size of the aperture of the zocecium. The distances 
of the zocecia from one another must not be measured in the 
basal portion of the zoarium, nor must we take the one or 
two lowest zocecia of an internode ; ; but, taking older zocecia, 
the variation is usually but slight, so that the typical size is 
easily obtained by measuring a few cases, and abnormalities 
through arrested growth are readily distinguished. 
Harmer has shown that the position of the basis rami give 
useful classificatory assistance. Some are, as Harmer says*, 
‘wedged in” between two zocecia, which occurs always in 
what I should call the denticulata seh) and in a few others, 
as C. conferta, B., C. s’gmoidea, sp. n., &e., but it is not 
common ; others have the basis ‘rami nar reaching to the 
next zocecium, as in some C. ramosa, Harm., and a large 
number of other species as seen in C. eburneo- rea (see 
Pl. XVI. fig.4). A third group has a very short basis rami, 
ending before it reaches the lower zocecium—as, for example, 
in C. eburnea (fig. 6), in some C. ramosa in which both long 
and “ graft” basis rami occur, and in a large number of the 
smaller forms of Crisia. This last group I have distinguished 
in my notes as the “graft”? form, as it reminded me of a 
grafted tree or shrub. “Tn connection with the basis rami 
there is still another character, perhaps of greater importance 
than the shape—that is, the width of the base of the inter- 
node, which may entirely differ from the diameter of the 
zocecia. ‘The bases rami that are wedged in are the widest, 
and I have elsewhere alluded to the fact that in most cases 
the size of the base of the lateral and main internodes is the 
same ; but in a few with “graft” basis rami, as C. ed- 
wardsiana, the base of the branches is slightly less than that 
of the main stem. In the denticulata group the basis rami is 
wider than in the other groups, being wider than a zocecium, 
as can be best seen in the new branch, where from the wide 
joint the first zocecium gradually becomes narrower. This is 
shown in my figure of what I considered was C. elongata TF 
and also in the figure now given of C. eburneo-denticulata, B. 
(fig. 4), the base of which is 0:12 mm.; the base of C. ser- 
* “On the British Species of Crista,” Micr. Journ. vol. xxxii. n. s., 
. 130 (1891). 
ine Proc Zool. Soe. 1914, pl. i. fig. 3. 
C. elongata was but sliat! tly described by Milne-Edwards, and a figure 
without a scale was given which really might pass for various species, 
