40 M. L. Cordier, Exammatton of recent Experiments 



According to this table, the depth, corresponding to the in- 

 crease of V Fahr. of heat, will be in round numbers as follows: 

 By two series of observations made during two years, in two 

 points of the mine of Beschert Gliick, from 101 to 68 feet, mean 

 84 feet; by four series of observations made in 1815, in four 

 points of the mine of Alte Hoffnung Gotes, from 175 to 64 feet, 

 mean 93 feet ; by two observations made cursorily in the United 

 Mines, 31 feet ; and by a series of obvServations which lasted 18 

 months, in a part of the Dolcoath mine, 54 feet. 



If we compare these numerical results with those previously 

 obtained, it will be seen that they lead to nearly the same con- 

 sequences. I shall therefore blend them in the conclusions 

 which are to be deduced from all that I have hitherto stated. 



But before resuming these conclusions, I must briefly state 

 several important considerations which may influence the judg- 

 ment that is to be formed respecting the merit of the experiments 

 which have been discussed. 



First Consideration. — It is not sufficiently known that ther- 

 mometers are almost always in some degree deficient in accuracy, 

 even when they come from the best artists. In the first place, 

 from the inadvertence of the manufacturer, the scale may have 

 been placed a little too high or too low. I have an instrument 

 of this kind, in which the zero of the scale was originally 0.3 

 'beneath the freezing point. In the second place, from the ef- 

 fect of very small inequalities in the bore of the tubes, differences 

 jof three or four tenths are very common in the progress of two 

 instruments regarded as tolerably good. I have often seen 

 greater variations. Lastly, through a vice inherent in the in- 

 strument, in proportion as it becomes old, the mercury keeps 

 higher than it ought to do, to correspond with the indications of 

 the scale, and this elevation, which is attributed to a slow con- 

 traction of the bulb, may sometimes exceed a degree. 



In making experiments on the subject which at present occu- 

 pies our attention, the first care of the observers should be to 

 verify the accuracy of the thermometers under these different 

 points of view ; and the second to give an account of the verifi- 

 cations. Unfortunately, the latter consideration has been ne- 

 glected by all observers^ so that, although it is to be pre- 



