368 REPORT— 1851. 



fig. 1, Plate XXII.) of the image of the surface of the mercury in the tube 

 '(at i') is rather more than twice the range of the mercury itself; that the com- 

 pensation was overdone by about -^gth of the wliole amount ; and that the 

 mean error of an observation, from a whole series of 181 comparisons, was 

 0*0028 inch. It was thought " only fair to suppose that a portion of this mean 

 error is to be ascribed to the standard barometer ;" and I fear a portion may 

 be attributed to dry rot or other cause of infirmity in the flooring of the 

 quadrant-room, that end of the frame-work (P) which carries the barometer 

 being much nearer to tlie wall of the building than the image end. "The 

 comparisons do in fact exliibit some symptoms of alterations in the floor," 



These comparisons, and the numerous computations, &c. required, were 

 made with great care, and yield a very satisfactory general result. "An error 

 of 0*002 inch in an observation would not be greater than might be expected 

 in a standard barometer*." 



Anemometer. 



On September 27, Mr.Howlet, of the Ordnance Office, brought an experi- 

 mental anemometer for comparison with our balance anemometer {vide p. 341, 

 ante), and for examination. 



It was speedily erected upon the leads at the southern end of the observa- 

 tory, and the action of both instruments was watched during about half an 

 hour. The mean result was, that the former gave about one-tenth more for 

 pressure of wind than the latter. 



At Mr. Howlet's request, a note was addressed to Sir John Burgoyne, in 

 which I expressed an opinion that an instrument of the kind, exhibited by 

 Mr. Howlet, might answer the purpose of a portable anemometer, applicable 

 in the intended manner, if properly constructed. 



Mr. Howlet made a sketch of his model, and left it at the Observatory. 



IV. MISCELLANEOUS MEMORANDA. 



On September 12, 1850, Mr. Jesse (of the Woods and Forests) visited 

 the building, and promised to represent its state as regards dry rot, &c. 



On October 5, an interesting account of an Aurora Borealis, which oc- 

 curred on the evening of the 2nd, was received from Mr. Gassiot, who, with 

 some friends, had carefully observed it at Clapham Common. The direction 

 in which it appeared was from N. by W. toN.N,W. bycompass, Mr. Gassiot's 

 letter was copied in our Diary. 



On November 5, Professor Graham came to confer about methods of 

 transporting air, free from town contamination, for chemical analysis, with 

 reference to the presence of ammonia, from this locality to the London Uni- 

 versity. Every means in our power, calculated to facilitate Dr. Graham's 

 important researches, were (of course) tendered with great alacrity and 

 pleasure f . 



* If the instrument were mounted in as substantial a manner as are the horizontal and 

 vertical-force magnetographs, and if a compensating apparatus of larger zinc rods and stout 

 glass tubes were substituted for this of small rods and wood, the error of an observation would 

 probably be less, and even the remaining error might perhaps be almost got rid of by making 

 ■use of the micrometer screw (at b^*), provided expressly for preventing the compensation from 

 being either " overdone " or underdone. The neglect of doing this is entirely my own. 



Since the above was vn-itten an instrument improved as above has been ordered by 

 Mr. Johnson for the RadcUfFe Observatory. 



t The aspirator (described at p. 344) would probably be found convenient in cases of 

 this kind. 



