638 Scientific Intelligence. 



carbon, the operation is conducted in the usual way until the 

 hydrogen has been oxidized, then the calcium chloride tube is 

 changed, and the residual carbon is burnt in moist oxygen. — 

 Berichte, xlvi, 949. h. l. w. 



3. The Heat of Combustion of Diamond and of Graphite. — 

 Roth and Wallasch have made new determinations of these 

 constants and give the following results : 



Diamond 7869 + 3 cal. per g. 



Graphite 7854 + 1 " " " 



The result for diamond agrees closely with that of Berthelot 

 and Petit, but these older investigators obtained a much higher 

 value, 7901-7902 cal., for blast furnace graphite. The new 

 result for graphite represents the higher values obtained by the 

 investigators from several samples of the natural substance and 

 the Acheson product. Other samples of natural graphite gave 

 results no higher than 7830-7835 cal., agreeing with remarkable 

 closeness with the result of Mixter, 7831 cal., published in this 

 Journal in June, 1905. It appears, therefore, that the heat of 

 combustion of diamond is higher than that of graphite, although 

 for a long time the opposite has been supposed to be the case. — 

 Berichte, xlvi, 896. h. l. w. 



4. Manual of Qualitative Analysis; by W. F. Hoyt. 12 

 mo., pp. 36. New York, 1913 (The Macmillan Company. Price 

 80 cents net). — This very small book gives very brief directions 

 for the detection of single elements and radicals by means of 

 reagents and in the dry way. Such a course, where separations 

 and interferences are left out of consideration might be very use- 

 ful to beginners, and it is to be remembered that Fresenius gave 

 such a course as a preliminary to more elaborate investigations in 

 his classical work on the subject. But, unfortunately, the book 

 under consideration contains many mistakes and misleading state- 

 ments. For instance, it is impossible to detect aluminum in the 

 manner described, by adding ammonia to the potassium hydroxide 

 solution ; cobalt sulphide is incorrectly assumed to be insoluble in 

 nitric acid, and thus to be distinguished from nickel sulphide ; in 

 testing for an ammonium salt with platinic chloride, the addition 

 of potassium hydroxide to the liquid is curiously specified. The 

 table of solubility at the end of the book has an astonishing num- 

 ber of mistakes. For instance, ammonium arsenate, ammonium 

 borate, and even ammonium carbonate are indicated as sparingly 

 soluble in acids. In this table the author gives the meaning of 

 the symbols "a" and "w" incorrectly as slightly soluble in acids 

 and water. He should have omitted the word "slightly," which 

 gives a meaning practically opposite to the true one. Taking 

 these errors with many others, there are probably more than 200 

 false statements in this single table of solubilities. h. l. w. 



5. A Course in General Chemistry ; by William McPhee- 

 son and William Edwaeds Hendeeson. Large 8 vo., pp. 556. 

 Boston 1913 (Ginn & Company). — The authors state that the 

 text presents few novelties either of arrangement or method. 



