ELISHA MITCHELL SCIENTIFIC SOCIETY. 73 



melting of the hydrate. The lowest temperature reached by blow- 

 ing was — 19.5"C., the relative humidity of the room being 75. When 

 the thermometer was wrapped with filter pajjer and this dipped in 

 disulphide, the reduction went on regularly to the constant point of 

 decomposition. The slow sinking at -9 C. was not noticed. 



Several observations showed that the reduction was in a measure 

 dependent upon the relative humidity of the surrounding atmos- 

 phere. This was determined at each observation of the point of 

 decomposition, but the exf)eriments could not be made accurately 

 enough to deduce any fixed law. Air saturated with moisture gave 

 no hydrate. From this point the temperature of decomposition 

 seemed to sink, the lowest points being reached when the air con- 

 tained between 70 and 80 per cent, of moisture. Below 60 per cent, 

 the temperature in a number of observations rose once more. Even 

 if the current of air was too slight to affect a small candle flame, 

 it would still show upon the delicate thermometer used, the varia- 

 tions caused being shown by shielding the bulb on one side or the 

 other. Entire surrounding and protection of the bulb interfered 

 with the evaporation, and of course the evaporation itself caused 

 currents. Some other way must then be devised for determining 

 this point. Guarding against draughts as carefully as possible, the 

 determinations made, some twenty-five or thirty in all, ranged from 

 — 14°C. to — 17^C. No point such as that mentioned by Wartlia 

 ( — 13=C.), to which the mercury would rise and remain constant, was 

 observed, the rise being rapid and regular till the temperature of 

 the room was reached. Berthelot gives the point of decomposition 

 as about — S'^C. 



Chemical Laboratory, U. N. C. 



NOTES ON THE INDIAN BURIAL MOUNDS OF 

 EASTERN NORTH CAROLINA. 



J. A. HOLMES. 



.So far as is known to me, no account of the Indian burial mounds, 

 which are to be found in portions of Eastern North Carolina, have, 

 as yet, been published. This fact is considered a sufficient reason 



10 



