REMARKS ON THE CLIMATE OF SOUTHAMPTON. 



55 



coincide with those at Greenwich. On frequent occasions I have applied 

 separately to the monthly means of the 9 a.m., 3 p.m. and 9 p.m. observa- 

 tions, the corrections given by Mr. Glaisher in the Phil. Trans, part 1, 1848, 

 and the results were in no case consistent : nor is this surprising, when we 

 regard the situation of Southampton at the head of an estuary which is di- 

 vided into two arms by the Isle of Wight. Most probably the local varia- 

 tions of atmospheric pressure are peculiar, but what these may be can only 

 be determined by a far more extensive series of observations than I have had 

 the leisure to undertake : one on the plan of photographic registration would 

 admirably answer the purpose. 



Mr. Glaisher on one occasion expressed himself to me unfavourably with 

 regard to barometers to which it was necessary to apply the capacity correc- 

 tion. His reason was, that as the mercury descends in the tube and the 

 cistern becomes fuller, a portion of the hollow cylinder of glass composing 

 the tube is enclosed, and the mercury rises higher than the capacity correc- 

 tion would indicate, by a quantity dependent on the volume of the section of 

 the tube vvhich it had enclosed. After having carefully considered the sub- 

 ject, I entered upon an investigation which should result in leading me to 

 reject the observations I had taken, or to confirm my confidence in them. 



Let a = the area of the hollow part of the 

 tube. 

 ra = the area of the surface of the cis- 

 tern. 

 h = the area of the annulus or section 



of the glass tube, 

 c = the ascent or descent of the mercury 



from the neutral point. 

 X = the correction required, + in the 



former case, — in the latter. 

 Then acz=(ra — h)x. 



• True height. 

 Supposed height. 



ra — b 

 Substituting the values of these quantities 

 measured from my barometer, the internal 

 diameter of the tube being 0'283 in., the ex- 

 ternal 0'41 in., and the proportion between 

 the area of the section of the tube and that of the cistern as 1 to 42, we 



have 

 then 



a = 0-0629, r -. 



42, b = -0691 ; let c = 1 inch, 

 •0629 



■2-535 — -069 



= •0248; 



but the correction for capacity, as applied in the usual way, 



^«£= 1-0238; 

 ra 42 



it follows therefore that when the mercury has ascended or descended from 

 the neutral point one inch, the diiference caused by the enclosure of the tube 

 will be exactly -001 ; a quantity less than that read by the vernier — proba- 

 bly less than the error of observation — and therefore one which may fairly 

 be neglected in practice. If therefore the exact proportion between the areas 

 of the tube and cistern be carefully -ascertained by the maker, we are able to 

 arrive at sufficient correctness with a barometer to whose readings it is neces- 

 sary to apply the capacity correction; the trouble of reading otf is less. I 

 have tabulated the capacity correction and the index error ; so that by taking 



