182 PROFESSOR THOMAS GRAHAM's SCIENTIFIC WORK. 



Essentially iuductive in his mode of thought, Mr. Graham developed 

 his leading ideas, one after another, directly from exjieriment, scarcely, 

 ifat all, from the prevailing ideas of the time. As well observed by 

 Dr. Angus Smith, " he seemed to feel his way by his work." His records 

 of work are usuall}', in a manner almost characteristic, i^receded each by 

 a statement of the interpretation or conclusion which he formed ; but 

 the records themselves are expressed in the most unbiased matter-of- 

 fact language. Singularly cautious in drawing his conclusions, he 

 announces them from the first with boldness, making no attempt to con- 

 vince, but leaving the reader to adopt them or not as he pleases. 

 Accordingly, in giving an account of his various researches, Mr. Gra- 

 ham rarely, if ever, deals with argument ; but he states succinctly the 

 experiments he has made, the conclusions he has himself drawn, and 

 not unfrequently the almost daring speculations and generalizations ou 

 which he has ventured. Some of these sj)eculations, on the constitution 

 of matter, are reproduced in his own words further on. 



Mr. Graham was elected a fellow of the Eoyal Society in 1837 ; cor- 

 responding member of the Institute of France in 1847 ; and doctor of 

 civil law of Oxford in 1855. 



The remaining pages of this abstract are devoted to an account of his 

 principal discoveries — the generalizations they suggested to him, and 

 the relations in which they stood to precedent knowledge. 



I. 



Modifications of pliosplioric acid. — At the date of Mr. Graham's inves- 

 tigation of this subject, when oxy-salts were usually represented as com- 

 pounds of anhydrous base with anhydrous acid, the point of greatest 

 importance, with regard to each class of salts, was held to be the ratio 

 borne by the oxygen of the base to the oxygen of the acid. Thus, in 

 the carbonates, this ratio was as 1 to 2 ; in the sulphates, as 1 to 3; and 

 in the nitrates, as 1 to 5. But with regard to the i)hosphates, taking- 

 common i)hosphate of soda as a type of phosphates in general, there 

 was a difliculty. Dr. Thomson maintained that, in this salt, the ratio 

 of the oxygen of the base to the oxygen of the acid was as 1 to 2; and 

 his view was substantially supported by Sir Humphrey Davy. Berzelius 

 contended, however, that the ratio was as 1 to 2i, or, to avoid the use 

 of fractions, as 2 to 5; but, notwithstanding the excellence of the 

 Swedish chemist's proof, and its corroboration by the researches of 

 others, the simpler and, as it seemed, more harmonious view of Dr. 

 Thomson prevailed very generally in this country. Anyhow, those 

 l)liosphates in which the oxygen ratio was the same as that in phosphate 

 of soda were taken as the neutral salts. But phosphate of soda was 

 found to have the peculiar and quite inexplicable property of reacting 

 with nitrate of silver to throw down, as a yellow precii)itate, a ])hosphate 

 of silver, in which the proportion of metallic base exceeded that in the 

 original phosi^hate of soda — the precipitation «f the basic salt being 



