2S2 F. !■:. WKKiTTT OUSTDIAN IMto.M 1(']':LAM) 



produced in cooling glass. These are wol1 known in the glass industry 

 and apply with equal force to the cooling of a silicate glass of nioldavitic 

 or rhyolitic composition, provided proper allowance ])e made for differ- 

 ences which arise because of high silica content. On the cooling of a 

 mass of glass heated to a high temperatui-e, tlie outer portions of the 

 mass in contact Avitli air cliill most rapidly and coiitract, Init on so doing- 

 meet with resistance from the hot interior mass. This slirinkage against 

 strong counter-resistance produces radial' compression in the marginal 

 shell, which, liecause of the raiiid cooling, quickly becomes so stiff that 

 appreciable movement is no longer possible ; the material thus sets under 

 a state of permanent radial compression. The central portion contracts, 

 in turn, on cooling and tends to draw away from the now rigid incasing 

 shell. Tensile stresses normal to the boundary surfaces are thus set up 

 and the material soon acquires a permanent set under tensile strain. 

 The net result of such rapid cooling is therefore an outer zone of radial 

 compression which decreases rapidly toward the center of the mass; it 

 becomes zero (neutral l)and or band of no strain) and passes finally into 

 a wide central region of tensile stresses. 



It is ol)\io\is that the relative intensities of the strain thus set up and 

 ihe relati\'e widtli^ <d' tlic zones of coiiipi-e^sioii and of dilatation depend 

 on the composition and size of the glass mass, on the initial temperature 

 of heating, and on tlie rate and conditions of cooling. Experiments have 

 shown thai in oi-dinary glasses the temperatui'c region at which the vis- 

 cosity of the material becomes so great that differential strains rtiay per- 

 sist for a period of time is between 2oO° and 4.")0° C. Above 500° prac- 

 tically all differential stresses are relie\c(l Ity Ibiw of the material, while 

 at 250° the movement in the material is so sluggish that a very long 

 period of annealing is i'e(|nire(l to ]n'oduce an apjn'eciable relief of sti'css 

 differences. At a still jow-ei- tempei-atni'e the glass is so rigid that under 

 small loads it beluncs as an elastic sdlid and the I'orces of restitnlioii set 

 up as a result of th(^ strain suffice to restore tlie material to its initial con- 

 figuration on I'elease of the load ; in short, strictly speaking, the glass is 

 no lonux'r \ iscons. accoi-ding to the established definition of the term. At 

 ordinary temperatni'cs the glass is so rigid, or its viscosity so great, thai 

 a state of strain may persist in it for geologic ages, as tests on obsidians 

 have shown. It is evident, therefore, that the state of strain of a glass 

 fragment may well serve as an indicator of the conditions under which it 

 cooled. 



The strain phenomena in glass are not apparent under ordinary con- 

 ditions of observation, Imt they can be rendered visible by simple optical 

 methods, Avhich in this respect function as does the developer on the pho- 



