12 Lieut. Col. Emmett's Meteorological Observations 



42° 4', leaving a discrepancy of 26', or there is no visible 

 light at that position indicated, instead of being of great 

 brightness. 



Again, for the corresponding radius of the secondary bow 

 we find 50° 17', and the difference between this and 42° 32' 

 is 1° 45', so that according to Mr. Airy the light should be 

 strongly visible to points distant 7° 45' from each other; whilst 

 the utmost limit to which the light can be traced is still distant 

 8° 30', and this difference leaves a strong presumption against 

 the theory, and induces a corresponding argument in favour 

 of the corpuscular theory, with which this fact is in ac- 

 cordance. 



III. Meteorological Observations taken at St. George^s, Ber^ 

 mudttf in the December half-year of IS"^*! ; introduced by Cor- 

 rections of Observations for the June half-year. By Lieut- 

 Col. Emmett, R.E.* 



MEAN height of barometer (Lond. & Edinb. Phil. Mag. 

 Nov. 1837.) should be for July 30-069, and not 30-161, 

 which is the height not corrected. The mean heights for Au- 

 gust and September should each be reduced -020, '054 having 

 accidentally been added for capillary action instead of '034. 



In the horary differences forwarded to Dr. Dalton it will be 

 observed the numbers of observations at the different hours 

 vary; they were also not always made on the same days: 

 but in revising my journal I selected those days only for 9 a.m. 

 and 4 p.m., on which the observations were made for both 

 hours ; and upon that revision the tables in your January 

 number were prepared. Such has been the course in the 

 half-yearly observations now inclosed, completing the year 

 1837. It is worth notice that the general fall of the baro- 

 meter from 1 P.M. to 4 p.m. usually and considerably exceeds 

 that from 9 a.m. to 1 p.m. Page 46, lines 4 and 5, should be 

 subtracted from the " former," not the latter. 



From a frequent comparison of the dew-point directly taJcen 

 with the wet-bulb thermometer I feel much confidence in it, 

 where suspended in a large room free from currents and fully 

 open to the air; but where exposed to currents, reflections from 

 the ground, &c. it cannot be fully relied upon, both it and the 

 register thermometer being affected thereby, and this it is often 

 difficult to avoid. As I before stated, the difference between 

 the wet and dry x by 2 and taken from the height of the dry, 

 gives the dew-point nearly, but rather too low, as the multiple 

 should be — 2 ; errors are therefore nearly doubled. 



* Communicated by the Author. 



