254 The Philippine Journal of Science ms 
Obviously these devices can be adjusted and used for washing 
with a flow of water in the opposite direction. For some 
classes of objects a large cylindrical vessel would serve the 
purpose as well as a conical one. The special advantage of the 
conical form is that it provides for the smaller, more slowly set- 
tling individuals among the specimens. Though prepared espe- 
cially for washing Volvox, the larger, conical, washing apparatus 
was successfully used for washing 4 cc of pollen of Cycas 
circinalis. 
METHODS OF ESTIMATING THE NUMBER OF CELLS IN 
SPHERICAL SURFACES 
For estimating the number of cells present in Volvox coenobia 
various methods have been employed. Those used by the earlier 
workers have been critically discussed by Klein. 4 
Leeuwenhoek counted the cells around the periphery of the 
spheroid and calculated the total number present from the count. 
Ehrenberg based his calculations on the same count, but, as 
Klein points out, his formula for the calculation was erroneous 
and gave numbers too large by more than half. Furthermore, 
his peripheral counts were, naturally enough, too large because 
of failure to limit the count to cells lying within an optical section 
no thicker than the average intercellular distance. The proper 
formula and coefficient for this method have been recently stated 
by Janet and are given below. 
Cohn 5 counted the number of cells present in a measured area 
of the spheroidal surface (100 microns square) and calculated 
therefrom the total number, deriving the area of the spheroid 
from the radius. A special variety of this method was used by 
the present writer, and is stated in detail below. 
Drude counted the cells in one-eighth of the spheroidal surface 
by means of a cross-hair ocular. 
Klein described the method which he found expeditious for 
making numerous estimates. Using a camera lucida, he drew 
the periphery of the coenobium and within it a group of cells not 
extending so far from the center as to give any considerable error 
due to foreshortening of the intercellular distances by projection. 
In this group of cells, drawn as points, he selected 4 to 6 as nearly 
as possible in a straight line. He then determined how many 
times the length of this line was contained in the periphery. 
4 Klein, L., Morphologische und biologische Studien ueber die Gattung 
Volvox. Pringsheim’s Jahrb. fiir wiss. Bot. 20 (1889) 145—146. 
5 Cohn, F., Die Entwickelungsgeschichte der Gattung Volvox, Festschrift 
(1875) 15. 
