Page:Proceedings of the Royal Society of London Vol 4.djvu/256

Rh maximum height of the barometer was 30°.89 in January and the minimum 28°.71 in October. The mean of the barometer at 9 was 29°.764, therm. 33°.455 ; at 2 29°.765, therm. 33°.327; and at 9 29°.784, therm. 29°.270. The diurnal observations would seem to support Professor Forbes's theory ; but the 9 observations are entirely opposed to it, as they appear with the same maximum sign as at the equator, whereas the sign ought to have been the reverse; indeed, with respect to the diurnal observations, the mean of five months of the year at 9 gives a plus sign, although the mean of the year at 2  only gives the trifling quantity of .001 plus. There is one remarkable feature in these observations that cannot fail to strike the meteorologist. M. Arago, from nine years' observations at Paris, reduced to the level of the sea, makes the annual mean height 29°.9546; twenty-one years' observations at Madras make it 29°.958; and three years' observations at Calcutta, by Mr. James Prinsep, make it 29°.764; and Mr. Thomas brings out 29°.771. That there should be this coincidence between the observations at Calcutta and Alten is curious. Neither Mr. Thomas nor Mr. Prinsep state whether or not their means are reduced to the level of the sea. It is to be suspected they are not.

For the next year, that is to say, from Oct. 1838 to Sept. 1839, both inclusive, My. Thomas uses a French barometer and French measurements, with centigrade thermometer attached to the barometer, and Fahrenheit's for the detached thermometer. He changes his time of observation from 9 to 8, 2 , and 8 , and he reduces his barometrical observations to centigrade. The results of the year are as follow : — mean annual pressure 29°.627 English; thermometer Fahr. 33°.36; greatest pressure in April, least in January!! The mean of 8 is 29°.620; therm. 33°.75. The mean of 2 is 29°.631 ; therm. 34°.73. And at 8 29°.631 ; therm. 30°.57. The diurnal observations assist to support Professor Forbes's theory; but as in the preceding year, the observation is at fault; and if the hour had been 9 o'clock instead of 8 o'clock, it would probably have been more so than it appears. The low annual mean state of the barometer for the year 1837-38 is even increased in the last year's observations; and as fresh instruments appear to have been used, there is ground to believe that the fact is associated with the locality, and it may be desirable not only to record in the Proceedings of the Royal Society the data already supplied, but to recommend to Mr. Thomas more particular inquiry on the subject.

The phænomena of the Aurora Borealis appear to have been observed by Mr. Thomas with great assiduity, and recorded with great care. On examining the register, with reference to M. Erman's important remark, that "in Siberia two kinds of aurora are distinguished, one having its centre in the west, and the other in the east, the latter being the more brilliant," it is found that twenty-two