PHYSICS. 347 



boiliug point be determined, and a long interval elapses before the zero 

 is determined, there is considerable error. If a thermometer is in any 

 particular molecular state, its reading will probably be in error, the 

 amount of which may be ascertained by placing it in ice and observing 

 the error of the zero reading. In order that a thermometer should 

 read correctly at any particular temperature, it should* bo exposed for a 

 considerable time to the temperature for which exact measure is desired 

 or else for a few minutes to a slightly higher temperature. {J. Phys., 

 December, 1S81, x, 520; Nature, July 28, 1881, xxiv, 294.) 



Tait has made an elaborate investigation of the errors caused by 

 pressure in the deep-sea thermometers of the Challenger expedition. 

 They were all registering thermometers of the Six pattern, the large 

 bulb being protected by an exterior shell of glass strong enough to re- 

 sist the i)ressure of at least 5,000 fathoms of sea water, or about six tons 

 weight per square inch, and filled with alcohol. The correction assigned 

 to them by Captain Davis, of the admiralty, was half a degree Fahren- 

 heit for every mile of depth. The first result reached was that this 

 correction needed was not due directly to the pressure, but probably to 

 the increased temperature produced by the compression. Calculation 

 showed that the interual capacity of a glass tube with thick walls is re- 

 duced by about one-thousandth i)art for each ton weight of pressure. 

 Hence if such a tube be partly filled with mercury with an index above 

 it, the index should be displaced by one-thousandth of the length of the 

 column of mercury for each ton weight of pressure applied to the out- 

 side of the tube. On testing the question with a thermometer tube, the 

 mercury column being a meter long, the index was found to be displaced 

 a millimeter for each ton of pressure. The apparatus employed for pro- 

 ducing the pressures under which the tests were made, of 11 or 12 tons 

 per square inch, is described, but the final results have not yet appeared. 

 {Nature, November, 1881, xxv, 90.) 



Waldo has examined with care three standard thermometers con- 

 structed for him at the Kew observatory. He concludes that between 

 0° and 100° C, the errors of these thermometers depending on the cali- 

 bration, are practically insensible. Direct examination of every degree 

 to detect accidental errors of graduation, requiring about 2,300 separate 

 micrometer readings, shows that no sensible accidental errors have been 

 introduced into the graduations of these standards. The corrections re- 

 quired at the freezing and the boiling point.'? were found to be, as a max- 

 imum, -i-0.38 at the freezing and -f-0.22 at the boiling point of a Fah- 

 renheit degree. {Am. J. Set., iii, xxi, 57, 1881.) Waldo has also sug- 

 gested two slight changes in the construction of the Kew standard 

 thermometers. As now made the capillary space is continued above 

 the calibrating chamber. As this causes serious inconvenience from 

 the lodgment of mercury in it, which is dislodged with great difficulty, 

 the suggestion is that the bulb extend to the end of the cavity. Since 

 it is often desirable to hang these thermometers up, it is convenient to 



