AT HIGH PEESSURES BY OPTICAL METHODS. 
129 
of silver chloride, these being protected from contact with the steel by lead paper. 
With both these arrangements pressures of up to 4,000 atmospheres have been 
reached without the glass being in any way injured. But in any case glass must be 
regarded as a very treacherous material when submitted to high pressure, and often 
a glass cone, which has withstood very high pressures, suddenly may crack at 
comparatively low pressures without having shown any signs of strain, and thus 
measurements are often suddenly interrupted. This circumstance adds, of course,, to 
the difficulties experienced in this kind of research at high pressures, and to the time 
required to carry through such research. 
The expenses for the construction of the above apparatus were defrayed by a grant 
made to the author from the “ Herman Rosenberg Fund ” of the University of 
Helsingfors. 
The preliiniuary experiments with glass windows of varying construction were 
carried out during the winter of 1910 in the laboratory of physical chemistry at 
Gottingen. Prof. G. Tammann kindly lent me a set of his pressure apparatus for 
several mouths, and thus enabled me to make the tests which were necessary before 
the apparatus described above could be constructed. I am also greatly indebted to 
Prof Tammann for many valuable suggestions, and for information on various points 
concerning the practice of high-pressure work. 
The apparatus has been mounted and brought into working condition in the Davy 
Faraday Research Laboratory of the Royal Institution of Great Britain, and the 
actual research work which forms the subject of Part H. of this paper has been 
carried out entirely in the Davy Faraday Laboratory. 
I have to acknowledge the liberality with which the Managers of the Royal 
Institution and the Director of the Laboratory, Prof. Sir James Dewar, have placed 
the technical resources of the Laboratory at my disposal. My thanks are also due to 
Sir James Dewar for the personal interest with which he has furthered the progress 
of the work. 
Part II.— Optical DETERmNATioN of Diagrams of State. 
4. General Methods. 
(l) Method of Ohserving the Substance. —In the case of the investigation of a body 
which remains liquid at all temperatures, and pressures at which the optical measure¬ 
ments are carried out, the substance may be allowed to fill up the entire interior 
space of the pressure bomb and the capillary tube through which the hydraulic 
pressure is transmitted. But, if the liquid has to be crystallized, the crystallization 
VOL. CCXII,—A. s 
