150 Analysis of ScietUific Books. 



searches on selenium. In the ninth section, on arsenic, he de- 

 duces the atomic weight of arsenic acid, from the experiments 

 of Berzelius and Laugier, to be 7.25; and, in the next page, 

 converts that number into 14.5, without trying to reconcile the 

 two statements. The atom of the metal itself, is, in the one 

 page, 4.75, and in the other, 9. In the first, it is said to take 

 Ij and 2^ atoms of oxygen, and in the second 3 and 5 to 

 produce its two acids. Tellurium and osmium, which are 

 classed with sulphur and arsenic, are re-printed with the addi- 

 tion of only one sentence to the first, on its seleniuret. 



This second genus of simple combustibles, the alkalifiable, 

 contains thirty-two bodies, which he subdivides into five fami- 

 lies. His long articles potassium and sodium are re-printed 

 verbatim, with the addition to the first, of a few lines from 

 Berzelius on the seleniuret. Lithia, is necessarily new ; but 

 the Doctor, as usual, accommodates its atomic weight to a 

 multiple of hydrogen. M. Vauquelin's estimate of its compo- 

 sition as a metallic oxide ; M. Arvedson's analysis of its chlo- 

 ride, and M. Gmelin's analysis of its carbonate, all concur in 

 shewing its atom to be 2.3, and not 2.25, as Doctor T. de- 

 cides. In the fourth section, which treats of calcium, he 

 makes the atomic weight of lime 3.625, by suppressing the 

 most accurate analyses of the carbonate by different chemists, 

 and trusting to his own, which was certainly incorrect, as the 

 sequel of his system shews. The sections on calcium, barium, 

 strontium, and magnesium, are literally re-priuted. His pre- 

 scription for preparing chloride of barium (muriate of barytes), 

 is marked with his usual awkwardness. By pouring muriatic 

 acid on the sulphuret obtained from the decomposition of the 

 sulphate with charcoal, he takes up all the iron, and contami- 

 nates the product unnecessarily ; whereas, by adding the acid 

 to the filtered solution of the sulphuret in hot water, this evil is 

 avoided. 



His second family, composed of yttrium, glucinum, alumi- 

 num, zirconium, and thorinum, are merely re-printed, and de- 

 serve no particular notice. The long article, iron, occupying 

 twenty-two pages, is re-printed with all its errors, on the for- 

 mation of steel, ^c, which have long been the ridicule of 

 practical men. The only thing new, is a short paragraph on 

 the seleniuret. Nickel and cobalt are also served up, as be- 

 fore. The absurdities exposed in our former critique *, on his 

 article manganese, remain unaltered in the present edition. 

 He has a new page on the hydrate of Arvedson, and the man- 

 ganesic acid of Chevillot and Edouards. Cerium, uranium, and 

 zinc, are re-printed verbatim. Cadmium is a transcript from 

 Gilbert's annals. His sections on tin, lead, bismuth, copper, 



* Journal of Science, VIII. 31?. 



