AT HIGH PRESSUKES BY OPTICAL METHODS. 
127 
hard viilcaii-fibre being substituted for the comparatively hard ivory. Quartz-glass, 
a heavy soft “ flint,” and a hard “ crown ” behaved in very much tlie same way as 
the borosilicate, but usually a cracking was noticeable v hen tlie glass was split up into 
lamellae, and this occurred in the case of flint glass, even at as low pressures as between 
200 and 300 kg./cm.b During these tests some very remarkable optical pbenomena 
were observed in the glasses in polarized light, Imt the desciiption of these, as well 
as of the details of preliminary work, cannot be entered on here, it being sufficient to 
state that it was found that the “ cleavage ” developed in the glass windows was 
due to tension and not compression of the glass cones, these being more rigid than 
the enveloping cone which flo'W's slightly, somewhat similarly to the way in which a 
rubber stopper flows and is stretched when it is forced into a bottle neck. The softer 
the material of the cone enveloping the glass the more marked this eflect was, and 
the sooner cracks were developed, and always at right angles to the axis of tlie glass 
cone, that is, at right angles to the direction in which the hydraulic pressure in the 
interior of the bomb acts. 
It was therefore thought that better results might be obtained by providing the 
glass windows with a steady support against rrhich they would be pressed, so that no 
part of the glass could be stretched as described alcove. But as it at the same time 
was essential to have a conical joint between the glass and the steel wall of the Ijomb, 
in order to secure a tight fit even at high pressures, the following way of fitting the 
glass windows was finally adopted :—The glass cone was inserted the opposite way 
into the conical place in the steel bolt, the smaller basis of the cone thus being acted 
upon by the pressure transmitting fluid in the interior of the l)oml), and the glass 
cone resting with its larger basis on a washer l)y which it is separated from its steel 
support. The steel and the washer supporting the glass cone are provided with a hole 
through the centre part, through which the observations are made, and the glass cone 
rests thus only with its outer part against the washer and steel. Tlie remaining 
circular space between the conical glass and the conical liore in the steel is wedge- 
shaped, and thus the material in it is forced in towards the narrower part, and, if 
softer than the glass, makes the joint tight, and transmits at tlie same time an 
all-sided pressure to tlie glass cone. This conical Avedge-shaped envelope has to be 
cut on the lathe exactly to the same angles as the conical surfaces of the glass and of 
the surrounding steel, the glass cone being finally ground in Avith A'ery fine emery to 
fit exactly. This composite cone, consisting of the glass and its envelope of “ fibre ” 
or elionite, is then ground into the conical space of the steel bomb, sufficient space 
being left for the AAmsher supporting the base of the glass cone. The arrangement is 
shown in fig. 4, Avhere a is the glass cone, h the conical AA^edge-shaped fibre eiiA'elope, 
c the washers, flat rings supporting the glass cone, d the steel-pressure bomb, e the 
steel bolt into which the glass windoAv is fitted, and f -du screAv. 
The chief object of these washers and conical packings may be described as being 
that of keeping the glass all the time surrounded by a half-plastic mass, Avhich fioAvs 
