Lieut. J. Herschel on Spectroscopic [June 17, 



Density of 

 Hydrogenium observed. 



When united with palladium 0-854 to 0*872 



"When united with palladium and platinum .... 0*7401 to 0*7545 



When united with palladium and gold 0*711 to 0*715 



When united with palladium and silver. 0*727 to 0'742 



The results, it will be observed, are most uniform with the compound 

 alloys, in which retraction is avoided, and they lie between 0*711 and 

 0*7545. It may be argued that hydrogenium is likely to be condensed 

 somewhat in combination, and that consequently the smallest number 

 (0*711) is likely to be the nearest to the truth. Eut the mean of the two 

 extreme numbers will probably be admitted as a more legitimate deduction 

 from the experiments on the compound alloys, and 0*733 be accepted 

 provisionally as the approximate density of hydrogenium. 



I have the pleasure to repeat my acknowledgments to Mr. W. C. 

 Roberts for his valuable assistance in this inquiry. 



Could the density of hydrogenium be more exactly determined, it would 

 be interesting to compare its atomic volume with the atomic volumes of 

 other metals. With the imperfect information we possess, one or two 

 points may be still worthy of notice. It will be observed that palladium is 

 16*78 times denser than hydrogenium taken as 0*733, and 17*3 times denser 

 than hydrogenium taken as 0*711. Hence, as the equivalent of palladium 

 is 106*5, the atomic volume of palladium is 6*342 times greater than the 

 atomic volume of hydrogenium having the first density mentioned, and 

 6*156 greater with the second density. To give an atomic volume to palla- 

 dium exactly six times greater than that of hydrogenium, the latter ele- 

 ment would require to have the density 0*693. 



Taking the density of hydrogenium at 0*7, and its atomic volume equal 

 to 1, then the following results may be deduced by calculation. The atomic 

 volume of lithium is found to be 0*826 ; or it is less even than that of hy- 

 drogenium (1). The atomic volume of iron is 5*026, of magnesium 

 4*827, of copper 4*976, of manganese 4*81, and of nickel 4*67. Of these 

 five metals, the atomic volume is nearly 5 times that of hydrogenium. Pal- 

 ladium has already appeared to be nearly 6 times. The atomic volume of 

 aluminium on the same scale is 7*39, of sodium 16*56, and of potassium 

 31*63. 



VII. " Spectroscopic Observations of the Sun" (continued) . By Lieut. 

 J. Herschel, in a Letter addressed to W. Huggins, F.R.S. 

 Communicated by Mr. Huggins. Received June 10, 1869. 



Bangalore, May 7, 1869. 

 My dear Sir, — After what I wrote to you last week you will scarcely 

 be surprised to hear again from me on the same subject ; and indeed I 

 feel in some measure bound to communicate without delay results of fur- 



