KELLICOTT: PERIODICITY IN ROOT OF ALLIUM 533 
The series of Podophyllum used as a check were treated in the 
same manner except that the smaller size of these tips necessitated 
their reduction to an average area of nine micrometer spaces. 
(4) ELONGATION 
In determining the periodicity of elongation the method of 
direct observation was used. Onions were grown in pine sawdust 
in pots with a narrow glass slip set in one side. Roots frequently 
grew almost vertically along this glass and would be exposed 
Properly to moisture and air in the pot. These bulbs were of 
course grown entirely in the dark-room. When the roots were 
about 35-40 mm. long they were observed by candle-light through 
a horizontal microscope and the temperature recorded, at one- or 
two-hour intervals during a 24-hour period. 
4. Daily periodicity of cell-division 
A. UNDER NORMAL CONDITIONS 
(2) Typical observation.—We shall consider first the rhythm of 
cell-division under normal circumstances of growth. To illustrate 
more fully the method employed and to furnish a basis for compari- 
son, one of the records has been reproduced completely in TABLE I. 
This table shows that by far the larger proportion of cells in 
the process of mitosis were in the prophase, as would be expected, 
since the preparation for and inauguration of mitosis occupy a 
much longer period than the later phases, which are passed through 
very rapidly. The table shows also how remarkably similar are 
the conditions of the root-tips examined at the same hour. This 
Similarity is much closer than was expected on account of the 
numerous factors liable to cause individual variations. It will be 
Seen that atg a. m., 11 a. m. and 1 p. m., for example, the num- 
bers of dividing cells are almost identical in the two tips, while 
only in a few instances, such as at 5 p.m., 9 p.m. or I a.m, 
were there differences at all considerable. 
The final averages of this table are represented in the form of 
a curve in FIGURE 2, curve J, which shows graphically the rate of 
cell-division during a 24-hour period. The curve is remarkably 
regular, with two maxima and two minima. Cell-division is most 
rapid (primary maximum) at II p. m., the secondary maximum 
