118 



CHEMISTRY. (CHEMICAL ANALYSIS.) 



raent which is electrically indifferent. The valency 

 of such an element would be null : it would, in this 

 direction, form the transition between the two suc- 

 cessive monovalent elements, the electro-negative 

 and the electro-positive. Assuming that the transi- 

 tion between the various series of elements as the 

 author has arranged tin-in in his treatise on the 

 systematic grouping of the elements is formed by 

 such elements, it follows directly that their atomic 

 weights ex pressed in whole numbers would be 4, 20, 

 86,84, 132. 212, ;iud 292, and the periodic system 

 would t hen contain 7 series of elements. With such 

 an arrangement of the elements the periodicity in 

 their properties appears as a continuous function. 



A Determination of the atomic weights of pra- 

 seodymium and neodymium by Harry C. Jones 

 gave for the former substance 140.47 when O = 16, 

 or 139.41 when O = 15.88: and for neodymium 143.6 

 when O = 16, or 142.52 when = 15.88. The 

 method of procedure adopted in the case of praseo- 

 dymium consisted in reducing the superoxide to 

 the sesquioxide in an atmosphere of hydrogen and 

 converting the sesquioxide into the sulphate. For 

 neodymium the oxalate was prepared from the 

 double nitrate of neodymium and ammonium, and 

 converted into the oxide by heating. Lanthanum, 

 which was present in the salts of both the metals, 

 was separated from them, but praseodymium in the 

 neodymium salt could not be got rid of, and had to 

 be calculated for. Absolute accuracy is not claimed 

 for either of the determinations, because of the im- 

 possibility of effecting complete separations of the 

 rare elements; but the author believes that his deter- 

 minations do not vary more than 0.2 from the cor- 

 rect atomic weights. His determinations differ 

 widely from those made by Von Welsbach, which 

 give 143.6 for praseodymium and 140.8 for neodym- 

 ium, (presumably) = 16; but the author observes 

 that if the values were reversed they would nearly 

 correspond, and suggests that there may have been 

 a typographical error in Von Welsbach's paper as 

 published. 



A. Rosenheim and P. Noge support the bivalence 

 of glucinum, because it forms with the binoxalates 

 compounds the preparation and properties of which 

 they describe, and which are ordinary double salts. 

 On the contrary, glucinum, like the bivalent metals, 

 forms with the bitartrates complex compounds 

 which are not double salts. The trivalent metals do 

 not form the double salts. The molecular weight 

 of chloride of glucinum, by the ebullition method, 

 gives a number corresponding to the formula 



(il< 'I 2 . 



B. Brauner has been led in the course of experi- 

 ments upon fractional crystallization to conclude 

 that cerium is associated with an element which 

 possibly has an atomic weight of 110; another earth 

 of lower atomic weight is perhaps present. From 

 experiments made upon ammonium thoroxalate, a 

 new salt which he describes, the same author de- 

 duces the atomic weight of thorium as Th = 232.44, 

 a result agreeing with the number obtained by 

 Kriiss and Nilson. Prof. Brauner has also contrib- 

 'itcii experimental data concerning praseodidymium 

 and ncodidymium. in the light of which he supposes 

 that the eighth series of the periodic system may 

 assume the form: 



Cs Ba La Ce Pr Nd 

 133 137.4 138.2 139.7 141 143.6. 



Starting from the chloride and using the purest 

 materials, F. P. Vonablc has calculated the atomic 

 weight of /irconium II = 1.008, O = 16, and Cl = 

 34.45 in three equations, viz.: maximum, 91.12; 

 mean. !i>.7* : minimum, 90.61. The atomic weight 

 as determined l>y Bailey is 90.95. The mean value 

 given in Clarke's recalculation is 90.40. The author 



purposes repeating the determinations with the 

 oxychloride. 



The equivalent of cyanogen has been determined 

 by G. Dean as 26.065 ; hence, if the atomic weight 

 of carbon be 12.01, that of nitrogen is 14.055. Mr. 

 Dean's method was by determining the amount of 

 potassium bromide that will react with a known 

 weight of silver cyanide dissolved in nitric acid. 



The atomic weight of boron is calculated by F. P. , 

 Armitage, from determination of the water of crys- 

 tallization of borax as 10. 



The atomic weight of cobalt has been revised by 

 T. W. Richards and E. P. Baxter, using the method 

 of the analysis of cobaltous bromide, and deter- 

 mined by them at 58.99, oxygen = 16. 



Chemical Analysis. The principal difficulty in 

 the analysis of silicates is the passage of the silica 

 through the gelatinous state. The evaporation to 

 dryness, to which recourse is ultimately had, tends 

 to the partial entanglement of the bases in the 

 residue. The analogies which exist between silicon, 

 titanium, and tin would lead us to suppose that it 

 is possible to obtain silica directly in the insoluble 

 state by attacking a silicate, which would not be- 

 come hydrated during its decomposition with suffi- 

 ciently concentrated nitric acid. A. Leclere hta 

 shown that this result can be obtained by previously 

 melting natural silicates with oxide of lead, a bas-e 

 which at a moderate temperature forms very fusi- 

 ble compounds with all the elements found as sili- 

 cates, and retains the alkalies in a well-marked and 

 characteristic manner. Porphyrized silicate is 

 mixed with oxide of lead and is heated for half an 

 hour or a little longer in a muffler at a reddish- 

 orange heat to fusion as a liquid enamel, which is 

 easily detached from the platinum of which the 

 crucible is made if the bottom is cooled quickly. 

 The enamel is decomposed by a mixture of not less 

 than ten times its weight of equal parts of ordi- 

 nary and of fuming nitric acid. By dilution with 

 boiling water the lead is dissolved while the silica is 

 collected on a filter. The insoluble hydrate ob- 

 tained by this method retains about 10 per cent, of 

 moisture when dried at 100 C. If we operate on^ 

 pure silica and prepare the enamel in sufficiently 

 thin sheets, so that there is no swelling under the 

 action of the acid, we can obtain plates that show 

 all the characteristic colors of the opal. The acid 

 liquor containing the nitrates is concentrated to get 

 rid of the excess of nitric acid, after which alcohol 

 is added. On the addition of hydrochloric acid, 

 slightly in excess, the lead precipitates immedi- 

 ately. 



The application of Nessler's solution for the de-L 

 tection of ammonia gas" in a gaseous atmosphere 

 has been found defective when amines of the fatty 

 series are present, since such substances deprive the 

 mercuro-potassic iodide in alkaline solution of its 

 special power. A more specific reaction for ammo- 

 nia is recommended by S. Deniges, which consists 

 in plunging the extremity of a glass stirring red, 

 wet with hypobromite of soda, into the gas to 

 be tested. On contact with ammonia gas the wet 

 part of the rod gives off a large number of sm.ill 

 bubbles of nitrogen gas, so small that they appear 

 as a white sheath round the extremity of the stirror. 

 At the same time the hypobromite loses its color. 

 The property of behaving in this manner to hypo- 

 bromite of soda belongs only to ammonia; the pri- 

 mary amines give a yellowish precipitate with this 

 reagent, while the other fatty amines cause no 

 notable phenomenon. M. Deniges describes two 

 other reactions which are less characteristic than 

 this one of the hypobromite of soda, but are inter- 

 esting on account of their sensitiveness. But. like 

 that of the mercurous nitrate, they belong as much 

 to the fatty amines as to ammonia gas. One is the 



