538 MM. Svanberg and Struve on the Atomic Weight 



terpoised with the crucible, and the wire left in the crucible. 

 The crucible with the residual alkali was heated to redness 

 and weighed to see how much had been taken for the expe- 

 riment. The crucible with the mixture of molybdic acid and 

 carbonated alkali was now exposed to a slowly-increased heat, 

 and as soon as the mass was perfectly liquid, the temperature 

 was raised as high as it was possible with an Argand lamp and 

 chimney. This heat was continued for half an hour, when 

 the crucible was allowed to cool over sulphuric acid ; the dif- 

 ference in weight gave the quantity of caibonic acid expelled. 

 This ignition, cooling and weighing, had to be repeated five 

 to six times with every experiment, and we took the average 

 of all these weighings as being nearest to truth. The reason 

 why we were compelled to take the average of the weighings 

 is, that no two successive weighings were found to agree, 

 there being sometimes an increase, at other times a decrease 

 in weight. We are not able to explain the phenomenon, but 

 it cannot be ascribed to the uncertainty of the balance, for this 

 does not amount to more than T Lth of a milligramme. 



Although these experiments have led to an atomic weight 

 for molybdenum which agrees very well with one arrived at 

 by another method, we cannot, on account of the variations in 

 the weighings, place perfect confidence in the experiments. 



Experiments with Chlorine Compounds. 



We hoped to arrive at good results by the preparation and 

 examination of the chlorine compounds of molybdenum ; in 

 this we have not succeeded, it being very difficult to obtain 

 any one of the chlorine compounds perfectly pure. Besides, 

 phenomena occurred which had not been previously observed 

 and investigated ; and, finally, we found it to be perfectly 

 impossible to determine the chlorine accurately in the form of 

 chloride of silver. 



Although for these reasons we have not followed up the 

 subject further, we shall describe our observations, as they 

 are not without interest. Oxide of molybdenum, prepared 

 by calcining a mixture of molybdate of potash with an excess 

 of chloride of ammonium and subsequent exhaustion with 

 water, had sometimes a brown, sometimes a gray metallic 

 colour, according to whether a larger or smaller quantity of 

 sal-ammoniac had been used ; in general it consisted of a mix- 

 ture of the two. This oxide was employed to prepare the 

 biacichloride of molybdenum. For this purpose we used a 

 glass tube furnished with two bulbs, about an inch apart ; into 

 one of the bulbs 4 grammes of the oxide were conveyed, and 

 it was then heated quite gently in a current of hydrogen, to 

 expel all humidity, and the tube connected with a chlorine 



