Supplement to "Nature" November 6, 1902. 



recognised. Pacific salmon received the attention offish- 

 breeders at a comparatively early date ; and, on account 

 of extensive fishing and the pollution of many of the 

 rivers, it is mainly owing to artificial propagation that 

 the supply of these fish is maintained on the western sea- 

 board of the United States. 



The description of the numerous forms of trout and 

 charr met with in the fresh waters of North America falls 

 to the lot of Mr. W. C. Harris, and constitutes (inclusive 

 of the angling notes) more than half the contents of the 

 volume. The author divides these fishes into salmon- 

 trout (commonly called brook-trout in America) and 

 charr-trout. That all the latter are specifically, if not 

 generically, distinct from the true salmon there can be 

 no doubt, although it has yet to be proved that this is the 

 case with the members of the former group. < >n this 

 point, however, the author is silent, although he admits 

 the extreme ■ difficulty of classifying these fishes in a 

 satisfactory manner. 



"The most prominent external marking by which the 

 salmon-trouts and charrs may be distinguished apart,'' 

 writes the author, " is the presence of red or crimson 

 spots on the body, the only exceptions being the great 

 lake trout, with greyish markings, and the Arctic trout 

 {Salvelim/s arcturus), upon which no reddish spots have 

 been observed." 



It was owing to the absence of these red markings that 

 the great lake trout, which now typifies the genus 

 Cristivomer, was formerly regarded as a true trout 

 instead of a charr. 



In addition to being a practical guide which should 

 be in the hands of every angler in American and 

 ( Canadian waters, this excellent little volume is a valuable 

 manual of North American Salmonida?. R. L. 



GAS ANALYSIS. 



Methods of Gas Analysis. ISy Dr. Walther Hempel. 

 Translated from the third German edition and con- 

 siderably enlarged by L. M. Dennis. Pp. xix + 490. 

 (London : Macmillan and Co., Ltd., 1902.) Price \os. 

 net. 



THE value of this well-known handbook on gas 

 analysis has been increased by additions both by 

 the author and translator, so much so that those who 

 already possess a copy of the first English edition will 

 probably consider it necessary to obtain also the present 

 one. The original work was practically restricted to 

 a description of operations which could be carried out 

 with the apparatus devised by the author, and this 

 character is still retained. The slight incompleteness 

 thus entailed is more than compensated for by the ex- 

 tremely practical nature of the instructions ; every process 

 described has been thoroughly tested and will work. 

 The author has found it advisable to abandon the division 

 into technical and exact gas analysis because, as he 

 stites in the preface, apparatus originally intended for 

 technical purposes may advantageously be employed for 

 many purely scientific investigations, and, on the other 

 hand, technical analyses must often satisfy the most 

 exacting conditions as to accuracy. The chief additions 

 to the first edition comprise"new methods for exact gas 

 NO. 1 723, VOL. 67J 



analysis and for the determination of combustible gases, 

 the separation of argon from the atmosphere,- improved 

 methods for the determination of carbon monoxide in 

 gas mixtures, the analysis of acetylene gas, the examin- 

 ation of gases produced by living bacteria, the simul- 

 taneous determination of fluorine and carbon dioxide, 

 the determination of the heating power of gases, the 

 estimation of sulphur in organic bodies and of carbon 

 in steel, and the analysis of the gases evolved in the 

 electrolysis of chlorides and the manufacture of bleaching 

 powder. The method originally adopted by the author 

 for the exact analysis of gases, although accurate, was 

 somewhat cumbersome to work and expensive to set up. 

 By adopting the principle of a compensation tube, 

 slightly modified from the suggestion of Pettersson, 

 the apparatus assumes a very practical form, gaining in 

 convenience and cost without loss of accuracy. The 

 determination of the heating value of gas, a determin- 

 ation which is rapidly increasing in importance on 

 account of the extended use of gas for heating and 

 power purposes and in the Welsbach incandescent 

 burners, has been usually carried out in calorimeters of 

 the Junker type. These are costly, require consider- 

 able amounts of gas, and must be carried to the place 

 where the gas is being used. In the ingenious apparatus 

 described by Prof. Hempel, a heating value can be deter- 

 mined on two litres, so that samples of gas can be brought 

 from a distance in metallic receivers and examined in 

 the laboratory. 



In the analysis of combustible gases, it is shown by the 

 translator that a modified Coquillion pipette, in which 

 the combustion is carried out by an electrically heated 

 platinum spiral, may in many cases advantageously 

 replace the usual explosion method. The error due to 

 the partial combustion of the nitrogen is avoided, and 

 owing to the use of oxygen instead of air much larger 

 quantities of gases can be burned with a corresponding 

 gain in accuracy, numerous test analyses being given in 

 proof of this point. The only suggestion which can be 

 made as to additions to this chapter is an investigation 

 as to the possible errors introduced into indirect explo- 

 sion analyses by the deviations of the various gases from 

 Boyle's law. According to Prof. Leduc, the errors from 

 this cause may amount in special cases to as much as 3 

 per cent, when the gases are measured at constant volume. 

 In the determination of carbon monoxide, a large amount 

 of space, some twelve pages, is devoted to a description 

 of the haemoglobin method, whilst the method of C. de 

 la Harpeand Reverdin, in which the monoxide is burnt 

 by contact with iodine pentoxide, is dismissed with a 

 short mention, although this method has been shown by 

 Nil loux, Gautier and others to be at least as sensitive as 

 the most refined modification of the blood reaction, and 

 is also applicable to coal gas. As it seems probable that 

 this method will supersede the doubtful cuprous chloiide 

 method, it would appear to have been worthy of a more 

 detailed examination. In this case, as in others in which 

 criticism might be offered, the author has preferred to 

 give prominence only to those methods with which he 

 has had personal experience. The work as a whole is a 

 most valuable addition to the very limited number of 

 works dealing with the handling and analysis of gases. 



G. N. H. 



