^20 THE POPULAR SCIENCE MONTHLY. 



glass, and an interior substance, still soft, plastic, and constantly 

 strained by a tendency to contract, to occupy smaller boundaries ; 

 but those boundaries cannot be moved without breaking. It is a 

 struggle of forces. If the thickness of the glass be considerable, the 

 constantly-increasing strain of contraction jjuUs so hard upon the 

 shell, that the force of cohesion is unable to withstand it, and the shell, 

 yielding with a shock, shivers the whole substance into fragments. 



In the process of annealing, the heat of the oven keeps the sur- 

 faces of the glass articles from absolutely becoming rigid, so that 

 they yield sufficiently to the strain of the contracting interior por- 

 tions; and if the whole substance of each article cools exactly to- 

 gether, the eltterior and interior all the time at the same temperature, 

 there is no strain and the ware is perfectly annealed. 



As it is practically impossible to accomplish a perfect equality of 

 temperature, a perfect equilibrium of the molecules cannot be ob- 

 tained ; but so near an approach to it is accomplished in a well-con- 

 structed annealing oven, that the cohesion of the glass is easily able 

 to withstand the trifling strain. 



In this view the action of cooling glass is simple and easily under- 

 stood, surely more simple than to imagine a tendency toward a fibrous 

 constitution of substance, or the imperious " cohesive polarity " of 

 the " Britannica " article. 



Test this theory upon the Rupert-drop phenomenon, and its expla- 

 nation will answer as well. 



A small amount of fliuid glass, when dropped into water, will im- 

 mediately, by the action of its heat, envelop itself in a garment of 

 steam, which protects its surface from contact with the water, until 

 that surface is so cooled that such contact fails to crack it. To test 

 this assumption, try the exj^eriment with partially cooled or soft glass, 

 and the result will be that all the drops will break in the water, on 

 account of cracked surfaces. With fluid glass, many drops will be 

 lost, not from the same cause, if the drops be not too large, but from 

 excessive contraction ; perhaps, out of a dozen only one or two will 

 be saved. 



The steam chills the surface of the glass much more rapidly than 

 the air does, consequently the inner and fluid glass in the Rupert 

 drop is inclosed in larger boundaries than if the drop had cooled in 

 the air. Hence contractive force is very strongly exerted to draw in 

 such excessive boundaries, but the curved form of the drop presents 

 arches of strength to aid the power of cohesion and resist destruction. 



One drop bursts in the water, another does the same, but perhaps 

 the third is drawn forth entire, though curled and twisted, as if in the 

 agony of its strain. Two of Nature's forces struggle fiercely for the 

 mastery in this little drop, that gives no indication of the contest as 

 it lies quietly before us. But break off" the thread, and down goes the 

 first of the little arches, that are holding up the surface against con- 



