September 2, 1904.] 



SCIENCE. 



309 



or the new description of the Morgan gems, 

 which is to present the finest combination of 

 realistic accuracy and artistic beauty yet at- 

 tained. George F. Kunz. 



Laboratory Exercises in Physical Chemistry. 



By Frederick H. Getman, Ph.D. New 



York, John Wiley & Sons. 1904. Pp. 241. 



Laboratory manuals in physics and in chem- 

 istry separately have been put upon the market 

 during the last twenty years in sufficient num- 

 ber to satisfy all reasonable demands on the 

 part of the general public. But during this 

 interval a field that overlaps both of these 

 has become differentiated, the start being made 

 by Ostwald, whose work has been taken up 

 and enlarged by a considerable number of 

 ardent workers. Many of the laboratory 

 operations involved are not provided for in 

 the current manuals in English. Ostwald's 

 'Physiko-Chemische Messungen' and Traube's 

 ' Physikalish-Cheraische Methode ' cover the 

 ground well in German, but, as is so often the 

 case in German books, the amount of detail 

 involved in the effort to be exhaustive, and 

 the large number of references to researches 

 not easily found in most American college 

 libraries, deprive them of much of their value 

 for American beginners. 



Dr. Getman's admirable little book has been 

 prepared with constant recognition of the 

 American demand for directness and economy. 

 His own experience during the last few years 

 in Johns Hopkins University, where physical 

 chemistry was the subject in which his doc- 

 tor's thesis was prepared, has been linked on 

 to several years of previous experience in the 

 teaching of chemistry. His effort has been to 

 select only such methods for presentation as 

 he has found to be typical and worthy of 

 preference. He has very decidedly the teach- 

 er's instinct, exhibiting much aptitude in the 

 art of arrangement and of clear expression. 

 Although the book is not yet two months out 

 of press, it has been already adopted in a num- 

 ber of university laboratories. It certainly 

 meets well the needs of the beginner in phys- 

 ical chemistry and is worthy of special com- 

 mendation as a handbook. 



The range covered may be briefly indicated. 



In the introductory chapters the author dis- 

 cusses the theory and use of the balance; 

 volume and density; viscosity and surface 

 tension; and the determination of solubility. 

 Thermometry and calorimetry are then con- 

 sidered, and a chapter on optical measure- 

 ments is introduced. This is followed by 

 several chapters on electrical measurement of 

 conductivity, electromotive force, current and 

 the dielectric constant. The last chapter is 

 on chemical kinetics as illustrated in reac- 

 tions of the first order, like the inversion of 

 cane sugar, and of the second order, like cer- 

 tain eases of saponification. 



The book closes with a well-selected series 

 of tables and an index. 



W. Le Conte Stevens. 



DISCUSSION AND CORRESPONDENCE. 



TPIE USE OF ROMAN NUMERALS. 



EoMAN numerals are frequently used to 

 designate the volume of a serial in biblio- 

 graphic references. Instead of writing Vol. 

 88, or merely 88 after the name of the serial, 

 we go to the trouble to write LXXXVIII. 

 Why ? Simply because we have seen others do 

 it, and have unreflectively imitated them. 

 When we are forced to defend our iisage we 

 find that there are few reasons for the use 

 of the Roman system, whereas there are many 

 reasons for the use of the Arabic system. 

 Those who are intelligently in favor of the 

 ' Boman numerals in bibliographic work argue 

 that the use of them enables us to avoid the 

 abbreviation for volume, while at the same 

 time we thus distinguish sharply between vol- 

 ume and part, or volume and page. They, 

 furthermore, urge that it is well for us to 

 conform to the usage of publishers. But 

 these arg-uments should be considered in the 

 light of the following facts. 



Although no one would deny that it takes 

 m\ich longer to write and read the Roman 

 numerals than the Arabic, and that we are 

 far more likely to make mistakes in dealing 

 with the former system, few of us realize how 

 great the difference in the ease and accuracy 

 with which we use the two systems really is. 

 In order that my arguments for the use of 

 Arabic instead of Roman numerals, not alone 



