168 CARNEGIE INSTITUTION OF WASHINGTON. 



cooling, nor carried into the middle of the melt; (2) that the glass shall not 

 become inhomogeneous through the precipitation of a crystalhne phase; 

 (3) that the bulk of the cooled melt shall be found cracked into large, reason- 

 ably rectangular blocks, having smooth, flat surfaces; and (4) that these 

 blocks shall be sufficiently free from strain to cleave readily with a smooth 

 fracture. 



The appearance of ream in the middle of the melt, vacuum bubbles, or a 

 crystalline phase, can be discouraged by rapid coohng, preferably from the 

 bottom of the pot, while the glass is still soft; and by insulating its top surface 

 as soon as the melt is set out of the melting-furnace. 



The cold melt shows cracks of two types — "spherical cracks," and "plane 

 cracks." The spherical cracks are due to decrease of temperature difference, 

 and first appear below the temperature at which the drop is a maximum; 

 these cracks form one after another until the temperature gradient has dis- 

 appeared. The plane cracks seem to be the result of a difference in rate of 

 contraction between the glass and the pot; the temperature at which they 

 first appear in a given melt seems to be nearly independent of the coohng 

 rate. The presence of either type of crack in the melt reduces the tendency 

 for the other type of crack to form. As the fracture due to spherical cracks is 

 rough and the pieces formed are irregular, while that due to plane cracks is 

 entirely satisfactory, it is desirable to maintain the temperature gradient at 

 a low value by slow cooling, and to prevent its decreasing much below its 

 maximum until after the formation of plane cracks has begun. This can be 

 accomplished by increasing the coohng rate at the proper time. 



The coohng rate can be reduced (1) by heating the surroundings of the melt, 

 i.e., placing it in a heated kiln; and (2) by surrounding it with an envelope of 

 some such insulating material as sand or kieselguhr. Variations of these two 

 methods are described and time-temperature data given. 



(27) Thermocouple installation in annealing kilns for optical glass. E. D. Williamson and 



H.S.Roberts. Bull. Am. Inst. Min. Met. Eng., 1445-1453 (1919). (Papers on 

 Optical Glass, No. 15.) 



During the war-time rush to prepare the glass necessary for the needs of 

 the Army and Navy, the problem of the temperature control of the annealing 

 kilns became most serious. This paper gives a short account of the system 

 evolved by the members of the Laboratory staff who were cooperating with 

 the Pittsburgh Plate Glass Company at its Charleroi plant. The points which 

 are most specifically treated are : 



(1) The advantages of the thermo-couple over other devices for this type 

 of work. 



(2) Choice of material for thermocouples. 



(3) Choice of measuring instrument. 



(4) Arrangement of leads and other apparatus. 



(5) The general problem of anneahng optical glass. 



The conclusions reached will interest and assist those who have to meet 

 similar problems in other fields. 



(28) Potentiometers for thermoelement work. Walter P. White. Bull. Am. Inst. Min. 



Met. Eng., 1763-1772 (1919). 



Thermo-couple pyrometers are read in three ways. First, by direct readers 

 where the current, and therefore the deflection, is proportional to the electro- 

 motive force of the couple; second, by potentiometers where the galvano- 

 meter merely helps to balance the electromotive force of the couple against 

 that of a standard cell by means of known resistances and a constant battery- 

 current; third, by intermediate instruments such as the pyrovolter, employing 



