510 ANNUAL RECORD OF SCIENCE AND INDUSTRY. 



of saltpetre, and eleven of Glauber's salt be dissolved in twen- 

 ty-two parts of water at 50 Fahr., the column of a mercurial 

 thermometer immersed in the mixture will fall to 4.10 Fahr., 

 or nearly the same as the preceding reduction. This, there- 

 fore, may be considered as much superior to any other com- 

 bination yet proposed for practical use in the production of 

 a low degree of temperature. The nitrate of ammonia has, 

 however, the advantage, even if more expensive, that it may 

 be used over and over again, it being only necessary to evap- 

 orate the solution to the point of crystallization, while the 

 mixture just referred to can only be used once. 



In one instance, with the air at 60 Fahr., and the water at 

 about 54, a thick, cylindrical cup of very hard ice, about 

 eight inches high and several lines thick, was produced in 

 about fifteen minutes. 



An interesting experiment bearing upon the same point 

 may be made by melting together fifty-nine parts of tin, one 

 hundred and three and a half of lead, and one hundred and 

 eighty-three of bismuth. If this be finely rasped or powder- 

 ed, and introduced into one hundred and eight parts, by 

 weight, of quicksilver, we shall find that the thermometer 

 immersed in the mixture will sink to 3.20 Fahr. ; and water 

 placed in a thin test-tube, and allowed to remain for a few 

 minutes in this bath, will be completely frozen. 15 C, xix., 

 301. 



A NOVEL SPERM CANDLE 



has recently been introduced in London. According to the 

 description, four lateral apertures near the lower end of the 

 candle communicate from the outside with internal longitu- 

 dinal passages, so as to admit air. By this arrangement, it is 

 claimed upward currents of air will be formed in the passages, 

 which, issuing in close proximity to the flame, will produce 

 more perfect combustion and increase the quantity of light 

 given out by the candle. The longitudinal passages may 

 either extend nearly up to the tip of the candle or may pass 

 directly through it. 



OBJECTIONS TO PRACTICAL USE OF OXYGEN IN ILLUMINATION. 



We have referred on several occasions to the use of oxvffen 

 as an illuminator on a large scale, and to the many advan- 



