457 
1910-11.] Isopiestic Expansibility of Water. 
however, a maximum limit of 1300 atmospheres ; they differ from all three 
in the sense that with very few exceptions the temperatures observed were 
also high, having a range from 200° C. to 1000° C. 
The dilatometer method used was such that while the whole dilatometer 
and contents were subjected to the high pressure, only that part of it 
containing the liquid under investigation was subjected to the high 
temperature. How this was done will be clear from the following. I shall 
first describe the compression-cylinder with its connections. I shall then 
discuss the various forms of dilatometer tried, and describe the general 
procedure with the form ultimately adopted. The last part of the paper 
will contain an account of the observations taken and the results obtained. 
Before the general description is begun, however, it should here be 
mentioned that most of the experiments were not made with pure water, 
but with a solution, slightly acid, formed by diluting 80 parts of a 
normal HC1 solution with approximately 1000 parts of pure distilled water. 
The main reason for this is that pure water attacks quartz-glass strongly, 
at high temperatures two or three experiments at most being possible with 
the same dilatometer bulb if pure water be employed. The HC1 solution, 
while differing very little from unity in density at room temperature, does 
not attack the quartz-glass so strongly. To justify the procedure several 
observations were made with pure water, and the results easily fell within 
the limits of error as determined by the other observations. They are 
included in the table on page 470 and are marked with an asterisk. When 
water is spoken of in the following, it is invariably this weak solution of 
HC1 that is meant. 
The Compression-Cylinder. 
As mentioned on page 456, 1 had at my disposal the compression-cylinder 
devised for higli-temperature work by Professor Des Coudres. By means of 
it one is enabled to subject a body to high pressures and high temperatures 
simultaneously. The general arrangement will be best explained by means 
of figures taken from the above-mentioned paper. B is a steel cylinder 
60 cm. long and 18 cm. diameter. It has an inner bore of 6 cm. width. 
It is closed above and below by plugs and K 2 , which are of the bayonet- 
joint pattern and rendered pressure-tight by vulcanite rings d,d. The steel 
tube z connects the compression-cylinder with a pump, compression-screw, 
and gauge as supplied by Schaeffer and Budenberg. During the course of 
the experiments the plug K a does not require to be removed. It carries the 
cooling spiral S, whose ends rr are hard soldered into it, and in which the 
water from the city mains is allowed to circulate. Through K x also passes 
