1867.] Physics. 281 



A somewhat interesting discussion has been going on in the 

 1 Chemical News ' for some time past, on the subject of Standard 

 Thermometers. Some of the facts elicited appear not to be known 

 so much as they deserve. It appears that the zero points of all 

 thermometers, as a rule, rise in a month or so after the instruments 

 are made. This rise varies generally between J° and 2°. The 

 bulbs of the best thermometers should therefore be blown some 

 months before the instruments are pointed. In this manner the 

 greater part of the error may be avoided. Even after all due 

 precautions have been taken, the thermometer should from time 

 to time be either compared with another standard which has 

 been repeatedly checked, or when this cannot be done, its zero 

 should be independently tested by means of melting ice. The use 

 of boiling-water is objectionable for the purpose of testing, as it 

 has a tendency to permanently raise the zero of the instrument, 

 even if it has been unchanged and correct before immersion. The 

 most likely cause of rise is the one-sided pressure of the air. The 

 bulb does not acquire, on cooling, its original size for some months. 

 Every thermometer loses its accuracy, for many months, whenever 

 it has served for the determination of higher temperatures ; and 

 there are very few thermometers in use in chemical laboratories 

 that do not come under this head. An instrument, after adjust- 

 ment, can only once be used for accurate determination of boiling- 

 points without readjustment — a circumstance always lost sight of 

 in chemical researches, and which explains, no doubt, many discre- 

 pancies between statements of different authors. 



We welcome, with pleasure, a very excellent text-book on the 

 subject of Heat.* A work of this sort has been long wanted for 

 the higher class of schools, and for students at college : and the 

 author, who is Examiner at two of our modern universities, where 

 physical science is directly encouraged, has had ample opportunities 

 of learning the wants of the students of the present day. The 

 book which he has now given to us is concise, without losing in 

 accuracy, or being obscure, and it embraces all subjects that could 

 be included under the question of heat, and treats of the latest 

 developments of the science, as discussed among physicists of the 

 present day. The discussion of perpetual motion, and the connec- 

 tion therewith of the questions of dissipation and conservation of 

 energy, may be mentioned as of especial interest. Throughout the 

 work theories give way to facts, the unknown is grouped markedly 

 from the known, and the methods of practical application of such 

 groupings of facts are fully explained by general formulae. With 

 this work, Professor Miller's ' Chemical Physics,' Professor 



* ' An Elementary Treatise on Heat.' By Balfour Stewart, LL.D., F.R.S. 

 Oxford : Clarendon Press. London : Macmillan & Co. 



