40
THE JOURNAL Off INDIAN BOTANX.
small and appeared unlikely to have any possible effect; on the water content this portion was neglected.
The size of the holes in the sieves varies with the thickness of the wire, so camera lucida drawings of a few meshes were made in each case under the microscope and carefully measured by comparison with drawings of a standard stage micrometer. In this way the average diameter of the holes was accurately measured, and this was found to eliminate some discrepancies noticed between results obtained with different sets of sieves. In the following tables, there- fore, we show only the average diameter of the holes and neglect references to the number of meshes to the linear inch.
 | A table should appear at this position in the text. See Help:Table for formatting instructions. |
As so large a proportion passed through the sieve of smaller mesh another set of three sieves was employed of 30, 60 and 90 meshes to the inch, with the following results
| A table should appear at this position in the text. See Help:Table for formatting instructions. |
It was obvious that the water content could not be explained in terms of these figures since the proportion of smallest particle in table 7 was in the order II, III, IV, I, and that of table 8 in the order II, III, I, IV, while the order of the water content was II, I III, IV. It seemed likely therefore that the water content must depend on the proportion of particles of a decidedly smaller size than those passing through a sieve of 90 meshes. No finer sieve being obtainable, and as it appeared doubtful whether it could be made sufficiently accurate in any ca%e, some other means had to be devised to solve the problem. ■ fvV/
