Tiri') ARSOni'TKlX AND l':MrssIOX OK All;. 17 



bou monoxide or not. The carbon dioxide itself seems not to pai'ticipate in the 

 photographic effect. For even tlie freshly tilled tubes gave diii'ing the first dis- 

 charges no other spectrum than the later photographs. Since the dissociation of 

 the carbon dioxide, which accoi-diiig to the general assumption precedes these oc- 

 currences, is dependent u[)on the height of the temperature, I have made prepara- 

 tions to institute experiments in which the heat of discharge shall be far below that 

 of the tubes I have hitherto iised. I hope, in this way, to obtain lietter information 

 than I have hitherto concerning the spectral Ijehavior of carbon dioxide. 



Carbon monoxide. — This concerns us here less as an ingredient of the atmos- 

 phere, which contains it in very small amount, than as one of those impurities of 

 Geissler tubes which have at all times created the greatest obstacle to the si)ectro- 

 scopic investigation of gases. Carbon monoxide was prepared : 1st, from oxalic and 

 sulphuric acids ; 2d, from yellow prussiate of potash and sulphuric acid ; 3d, from 

 formic and sulphuric acids ; and it was then washed with caustic potash solution 

 to retain the sulphui- dioxide or the formic acid, as the case might be, as well as 

 traces of carbon dioxide. 



Carbon monoxide absorbs the most refrangible rays somewhat less than carbon 

 dioxide, and gives, like oxygen, a series of rhythmical, inverted groups of lines. 

 The clearness and sharpness of these lines ai'e less than with oxygen, but far more 

 perfect than with carbon dioxide. 



The emission-spectrum has already been sufficiently considered in treating of 

 that of carbonic dioxide, so that I have nothing to add here upon the subject. 



Aqueotis vapor. — The filling of the Geissler tube was effected from a small 

 o-lass vessel with a faucet containing a few drops of distilled water. This vessel 

 was melted on laterally to the tube connecting the (Teissier tube with the air-pump. 

 Aqueous vapor was evolved in the absorption-tube by introducing into it before 

 exhaustion a little cup containing a few drops of water. Then, by pumping out 

 the tube repeatedly, it was at last filled with aqueous vapor free from aii-. Un- 

 fortunately this process has not, thus far, led to certain results, in consequence of 

 the deposit of dew. I must, therefoi'e, reserve a report upon this mattei- until these 

 experiments have been repeated. What takes place in the tube of aqueous vapor 

 is not deal-. The spectrum differs according to the mode of discharge. It consists 

 mainly of hydrogen lines, of the oxygen maximum of 185 MM, and of a great num- 

 ber of other lines, concerning \vhich I have not yet been able to make quite sure 

 whether they belong to aqueous vapor or not. Upon these spectra is laid, in addi- 

 tion, the more or less intense spectrum of carbon monoxide. Finally, by a suitable 

 choice of pressure, form of tube, strength of current, and mode of discharge, all 

 these spectra can be greatly weakened and the spectrum of the electrodes of the 

 tube brouc^ht out. My numerous observations indicate a regular dissocuition of 

 the aqueous vapor into" its elements accompanied by a simultaneous recombination 

 to form water. I have been led to this conclusion more paiticularly by the study 

 of the spectra of certain Geissler tubes which carried an absorption vessel or drymg 

 vessel as the case might be, which was cut off from the proper place of discharge 

 by a mercury valve, the arrangement being such that the vessel could be opened 



