150 Scientific Intelligence. 



inside the box, gave the temperature at any given time, and the 

 range of variation of temperature was recorded by two maximum 

 and minimum thermometers, fixed inside to the roof of the inner 

 box. To obtain ice free from air, water was boiled and then 

 frozen. It was then melted and again frozen in a mould. Some 

 difficulty was found in holding the ice-bars in the testing machine. 

 The mode which answered best was to freeze the ends of the ice 

 bar into conical metal collars, which fitted the shackles of the 

 machine. Extensions were measured by vernier callipers reading 

 to one-fiftieth of a millimeter between marked points on the metal 

 collars. To determine if any appreciable effect was due to dis- 

 tortion of the enlarged ends of the bars in the metal collars, 

 pieces of paper were gummed on the ice, and the extensions also 

 measured between fine pencil marks on these pieces of paper. It 

 was found that nearly all the stretching observed in measuring 

 between the metal collars was due to stretching of the bar of ice, 

 and only a very small part to shearing action in the collars. In 

 consequence of rapid evaporation from the surface of the ice bar, 

 the stress with a fixed load on the lever increased from day to day. 



Three experiments are given on bars initially about 234mm. in 

 length, loaded to stresses of from 4*3 to 2*0 kilos, per square cm., 

 and lasting from four to nine days. 



The three experiments show that ice subjected to tension 

 stretches continuously by amounts which depend on the tempera- 

 ture and the tensile stress. When the stress is great and the 

 temperature not very low, there are extensions amounting to one 

 per cent, of the length per day. So continuous and definite is 

 the extension, that it can even be measured from hour to hour. 

 These extensions took place at temperatures which preclude the 

 possibility of melting and regelation. 



The author hopes that on resuming the experiments next winter 

 at St. Moritz, he may be able to determine more exactly the law 

 of the extension. He has shown already that the extension in- 

 creases continuously with all stresses above one kilo, per square 

 cm., and at all temperatures between — 6° C. and freezing. 

 When ice is in a condition such that the point of a needle will 

 cause a set of radiating fractures to pass from the point of con- 

 tact in all directions, it stretches as certainly, though not by so 

 great an amount, as when it will permit the passage through it 

 of the same needle without showing flaw or scar. 



In the first experiment there was a total extension of 11 mm. 

 in nine days ; in the second of 1*8 mm. in five days ; in the third 

 of Vj mm. in three days. If we assume the extension propor- 

 tional to the time, there was a mean daily extension of 1*2 mm., 

 0*36 mm., and 0-56 mm. respectively. The stress in No. 1 was 

 greater than in Nos. 2 and 3, and the temperature not so very 

 low in the day, though low at night. In No. 3 there was a low 

 stress, but comparatively high temperature. — Proc. Hoy. Soc, 

 xlii, 331. 



3. Steam Calorimeter. — A remarkable, simple and exact method 

 of determining the specific heat of a solid has been perfected by 



