92 The Botanical Gazette. [March, 
ping of water close against the central zinc tube. Here the 
centrifugal force is slight and the water has opportunity to 
soak in while slowly moving to the periphery. After a short 
time the water supply can be so adjusted as to compensate 
evaporation without sufficient excess to be thrown out around 
the side of the pan. : 
Seedlings of peas, beans and corn have been grown in saw~ 
dust in this manner for several days, revolving constantly at 
a rate of 180 revolutions per minute. 
HORIZONTAL CLINOSTAT. Plate IX, figure 3, and details 
e, f, g, h, t, 7, B, 1, m, of plate X.—A wooden shaft, ¢, 1S 
turned 303 inches long and 13 inches in diameter, after a steel 
core has been inserted and then clamped in the lathe chuck 
in the manner previously described. A pulley, g, 23 inches 
in diameter, serving also as a shoulder for the larger pulley 
and table of the clinostat, %, is turned on the shaft in ae 
manner previously described so that its upper face stands z 
of an inch below the upper end of the shaft. The upper face 
is hollowed slightly. The table of the clinostat is made © 
the same size and in the same manner as that of the centrif- 
ugal machine, except that the central opening, 1} inches 10 
diameter, extends only 3 of an inch deep from the lower sur- 
face. This is crowded on the shaft against the small pulley 
and screwed to it. The core in the lower end of the shaft re- 
volves in an iron bearing similar to the one described for the 
centrifugal machine. The bearing for the upper end of the 
shaft, shown in figure 7 of plate X, is made in the following 
manner: a piece is cut from Z inch stuff 3 inches wide and 27 
inches long. Two inches from one end a hole is bored of the 
diameter of the shaft and a block is cut out through the 
middle of the hole as shown in the figure. When this is re- 
placed with long screws it serves as an adjustable box for the 
shaft. To set up the clinostat, the last mentioned piece 'S 
laid across the table with the box end projecting inwards. 
Another piece seven-eighths inch by three inches by twenty- 
two inches is placed beneath the table and the two pieces are 
then clamped against the table by means of a bolt and thumb- 
screw, &, at each end of the two pieces. This method of 
clamping apparatus to the table is easy of execution and 
insures rigidity; it also affords a simple method of tightening 
belts by loosening the thumbscrew and slipping the app@t- 
atus along the table. The iron cup which receives the core 
