35 ,q:; 



NOTES ON SOME REMARKABLE ERRORS SHOWN BY 



THERMOMETERS. 



By H. C. Eussell, B.A., F.E.A.S., G-overnment Astronomer. 



{Mead before the Moyal Society of N.S.W., 7 June, 1876,] 



In the present day, when so much reliance is placed on ther- 

 mometers both by scientific and medical men, probably no apology 

 is necessary for bringing before the members of the Eoyal Society 

 the faults of one or two instruments, when, as in the present 

 case, those faults seem to be quite inexplicable by known con- 

 ditions affecting the accuracy of thermometers, and to depend 

 upon some unknown relation existing between the mercury and 

 the glass. 



I therefore put the following facts on record, in the hope 

 that they may yet be found to be connected with the explana- 

 tion of some of the extraordinary temperatures that have beeji 

 published. 



T^or more than five years we have had a first-class dry and 

 wet bulb hygrometer in use at the Observatory ; by the side of 

 the dry bulb a standard thermometer has been kept, and always 

 read at the same time as the dry bulb. The difierence between them 

 varied very little, two or three tenths of a degree usually, and in 

 some rare cases as much as one degree. Up to the 26th February, 

 this year, we never had reason to suspect the dry bulb of uncertain 

 indications ; on that day the maximum temperature rose to 96"4 

 at or about noon ; at 3 p.m., the dry bulb and standard had both 

 fallen to 83"7, and at 9 p.m. to 68"9 and 69'0 ; next morning they 

 read 69 6 and 69*8, but the following morning the readings were — 

 standard, 64*9 ; dry bulb, 87*3 ; showing a difi'erence of 22*4. It 

 was at once inferred that the glass had cracked and let in the air, 

 but as no crack could be seen on careful examination, and it was 

 determined to continue observing it for a time. The observations 

 will be found in Appendix A. 



If a glass thermometer cracks, the mercury steadily rises till 

 the tube is full, and it was expected that such would be the case 

 here ; but no, the diiference steadily decreased, and in 35 days 

 it had almost recovered its original condition, being sometimes 

 less than half a degree ; between the 7th and 17th of April the 



D 



