396 Notices respecting New Books. 



effects conld be obtained when a suitable transmitter is placed 

 in the circuit, articulation being clear. 



27. The very beautiful experiments of Prof. Hughes and 

 of Prof. Ewing on transitory currents were made on wires 

 in which torsional " set " had not been produced. In the 

 experiments which have been described, rods with permanent 

 torsional sets were always used, except in experiments (14) 

 and (16). 



L. Notices respecting New Boohs. 



A Course of Experiments in Physical Measurement. — Part III- 



Principles and Methods. By Haeold Whiting, Ph.D. Boston, 



U.S.A. : D. 0. Heath and Co. 

 ~TN the two volumes of this work which have already appeared 

 *- the author has described the processes used in the measurement 

 of Physical Quantities, and has given full details concerning sources 

 of error and the various methods of avoiding or eliminating them. 

 Those parts of the theory of the subject upon which the experi- 

 ments previously described are based are now briefly treated of in 

 the book before us. The theory of errors also receives a very large 

 share of attention, and in connexion with it is included the calcu- 

 lation of the mean from a series of observations, and the estimation 

 of the probable error in the result. 



The latter half of the volume consists of a collection of mathe- 

 matical and physical tables, together with notes explaining their 

 use. In the case of the tables of powers, roots, and reciprocals of 

 numbers, and in one or two other cases also, two tables have been 

 given ; a rough one for all numbers of less than three digits, and 

 a more complete one accurate to five figures. The former should 

 prove useful for rapid work, and the latter for more refined calcu- 

 lations. We notice, however, that two tables are omitted which 

 we have found to be of immense value in ordinary calculations 3 

 namely a table of four-figure logarithms and a corresponding one 

 of antilogarithms. To be of maximum utility these should be 

 placed in a conspicuous position, say at the end of the book, and 

 should be so arranged that the whole of the logarithm table can be 

 seen at one opening, and the same in the case of the antilogarithms. 

 In our experience such tables are quite accurate enough for almost 

 all calculations with the exception of delicate weighings on the 

 balance, where the five-figure logarithms given by the author would 

 be required. There is moreover a positive objection to the use of 

 too elaborate mathematical tables, since they deceive the student 

 into thinking that his results have the same order of accuracy as 

 the tables. 



The Physical Tables are fairly complete and accurate. The only 

 error we have noticed is in the table of electromotive forces of 

 cells, where the positive pole of the Clark cell is erroneously stated 

 to be carbon instead of mercury. James L. Howard. 



