532 
Reginald Ruggles Gates 
Iu all tlie tissues measurements were taken in two directions in 
tke plane of section, giving the lengtli and tke breadtk of tke cell. 
In all the tissues examined except the pollen mother cells, the cells 
were usually roughly reetangular iu cross section so tliat the direction 
of measurement was thus determined. The pollen mother cells are 
usually nearly isodiametric or roughlv spherical in outline, but fre- 
quently one diameter is greater than the other, so that two measure- 
ments were also taken of these, one being the long diameter, and 
the other the diameter approximately at right angles to this, wliich 
is usually the shortest diameter. An average of these two was taken 
to represent approximately the average diameter of the mother cells. 
The margin of error is prohably larger in these than in the rectangu- 
lar epidermal cells. 
In another series of measurements of the mother cells only one 
dimension, the greatest diameter, was taken. Reference to the table 
will show that this series agrees very closely with the other. In 
most cases two series of measurements, made at widely separated 
tiines, are given, and they are nearly all consistent with each other, 
indicating that the numbers of measurements are large enough to 
give fairly accurate and reliable data. The records of individual 
measurements, from which this table is constructed, are preserved for 
reference and could be used in plotting curves of variability, etc. 
But a Statement of the ränge of variability in each case, as is 
given in the table, is all that is necessary in the present Connection 
(Table I). 
An examination of the table brings out many interesting facts. 
In the first place in every tissue compared the average di- 
mensions and volume of the cells is larger in 0. gigas 
than in 0. Lamarckiana. But further, the amount of difference 
varies markedly for different tissues. Some apparently following 
Boveki’s law and others departing rather widely from it. The rela- 
tive volumes of the cells were not computed until the observations 
and measurements were completed, so that no preconceived idea 
could have come in to affect the results in this regard. It will be 
seen, bowever, that most of the ratios (Table II) approach remarkably 
near to whole numbers or simple fractions such as y 2 or 2 / 3 . Whether 
these fractions have any particular meaning cannot be stated at the 
present time. 
In computing the average volume of the cells from their aver- 
age dimensions two formulae were used. For rectangular cells, such 
