Notices respecting New Books. 457 



from which, as is well known, ammonia is not liberable by treat- 

 ment of the salt with potash, or by its desiccation at upwards of 



100°. The base N 2 H 5 ^ though not procurable in the free 



state, as upon the above view of the cause of its stability it 

 scarcely should be, is yet transferable from one salt to another 

 by double decomposition with almost as much facility as am- 

 monia itself. 



What I conceive to be the constitution of the different plati- 

 nous and platinic ammonia compounds in relation to each other, 

 is indicated in the last chapter of my ' Outlines of Chemistry/ 

 just published. 



It is observable that in no stable metallicized ammonia hydro- 

 chloride is the number of nitrogen atoms more than double the 

 number of chlorine atoms in the salt. Thus the empirical for- 

 mulae of the purpuro-cobaltic and luteo-cobaltic chlorides are 

 Co 2 CI 6 , 10NH 3 , and Co 2 CI 6 , 12NH 3 respectively. These ex- 

 pressions are of course easily translatable into forms harmoni- 

 zing with the above suggested view of the constitution of con- 

 densed ammonia compounds. 



LVI. Notices respecting New Booh. 



Methods of teaching Arithmetic. A Lecture addressed to the London 



Association of Schoolmistresses. By J. G. Fitch, M.A. Pp. 31. 



London, 1869. 

 The School Arithmetic. By J. Cornwell, Ph.D., and J '. G. Fitch, 



M.A. Pp. 144. Tenth edition. London, 1869. 

 The Science of Arithmetic. By J. Cornwell, Ph.D., and J. G. Fitch, 



M.A. Twelfth edition. Pp. 372. London, 1868. 



\JITE have put these books together at the head of a short notice 

 * " on account of their common authorship, and of their being 

 more or less supplementary to each other. The first of them (the 

 lecture on methods of teaching arithmetic) contains many hints and 

 remarks likely to be useful to the audience to which it was addressed. 

 The point most dwelt on is the need of making learners understand 

 the ultimate reasons of the rules for performing the elementary ope- 

 rations of arithmetic, such as the rules for multiplication and division 

 of integers. We doubt whether the importance of this point is not 

 somewhat exaggerated. Any ordinary child of nine or ten years 

 can be brought to divide, for instance, 5382 by 23 correctly, and be 

 made to understand what is meant by the answer, viz. that if 5382 

 marbles were divided equally between 23 boys, each boy would get 

 234 marbles. But to make the child understand each separate step 

 of the process of the division is quite another matter. And though 

 much can be done by a good teacher by means of a discussion of 

 particular examples, yet we question whether any but a few ex- 

 Phil. Mag. S. 4. Vol. 38. No. 257. Dec. 1869. 2 H 



