- 688 PROCEEDINGS OF SECTION E. 
number of samples to be dealt with in a short time. Experi- 
ments were also tried with a modified type of the Nicholson 
hydrometer, the fiducial point being a glass pointer projecting 
upwards from the hydrometer bulb instead of a mark on the stem 
as usual. The contact of this pointer with its reflection at the 
surface of the water, as seen by looking upwards through the 
hydrometer glass, could be observed w ith very great precision, 
the whole arrangement being similar to that of the fiducial point 
of a Fortin barometer inverted. The ease with which separate 
determinations could be obtained by Mr. Buchanan’s method, using 
different weights, caused the writer to finally adopt it. The 
hydrometers used were two in number, the first having a volume 
of 904 cubic centimetres with stem diameter of, approximately, 
5 millimetres, and the second a volume of 1374 cubic centimetres, 
with stem diameter of, approximately, 4 millimetres. Both instru- 
ments were made by a Hicks of London. The diameters of the 
stems were carefully measured by micrometer gauges and the 
distances between the graduation marks verified on arrival of the 
instruments in Australia, and the accuracy of the zero graduation 
verified by testing in distilled water. From the above informa- 
tion and the weights in vacuo cf the hydrometers and weights 
used (which were carefully obtained), the value of each stem- 
reading of the hydrometer for each applied weight was ascertained. 
These were checked by testing the instruments in solutions of 
known densities. The hydrometers are standard at 60° Fahr., 
and the corrections to apply to readings taken at other temper- 
atures were carefully determined by experiment. The densities 
were determined, as far as possible, when the surrounding temper- 
ature did not differ greatly from 60 degrees. 
The observed densities are given in the attached tables, A to F 
inclusive, and are means of three readings of the hydrometer with 
different weights in each case. In almost every case the results 
obtained with the large hydrometer for each individual reading 
does not differ from the mean by more than 0:0006, the average 
difference being about half of this amount. With the small 
hydrometer these figures would be doubled. The densities given 
throughout are those of the sea-water at 60° Fahr. referred to dis- 
tilled water at 39:2° Fahr. as standard. The temperatures given 
are those reported by the gentlemen obtaining the samples, all of 
whom were carefully instructed as to the importance of reading 
the thermometer whilst the bulb was in the water. The ther- 
mometers used were ordinary ship’s thermometers, the special 
thermometers which the writer had ordered not being available in 
time. Attempts were made to obtain the samples in the months of 
February, May, August, and November, as being months when 
the greatest variations (ze., in Iebruary and August), and mean 
results (z.e., in May and November) might be expected. 
