108 Experimental Examination of the so-called Storm-glass. 



Mar. 13. For some days the lower deposit has been nearly 

 equal all round ; but a few days of cold weather heaped it up 

 against the side nearest the window. This afternoon it was 

 hollowed out : as soon as sun left window, the top thermometer 

 sank about two degrees, and the stars set in in currents. 



Mar. 29. 



Sun on. 4.40 p.m. Upper, 72. Lower, 69. Outside, 62. 

 Sun off. 5.40 p.m. „ 66. „ 66. „ 58. 



Mar. 30. 8.40 a.m. „ 54. „ 53. „ 52. 



Register, 47° ; feathers fine and falling. 



6.0 p.m. Upper, 56°. Lower, 55°. Outside, 53°. 

 Feathers softening down by solution. 



April 1. 8.0 a.m. Upper, 42°. Lower, 39°. Outside, 39°. 

 Register, 31°. Feathers sharp. 



I think it may fairly be concluded from these experiments and 

 observations that the storm-glass acts as a rude kind of thermo- 

 scope, inferior, for most of the purposes of observation, to the 

 thermometer. It does not seem to be capable of reference to a 

 standard, and hence observations made with it scarcely admit of 

 being registered, although attempts at a scale are made by some 

 instrument-makers. If, however, two or more of such gra- 

 duated instruments be placed in and about a house, their indi- 

 cations will vary considerably, according as they are more or less 

 exposed to the action of radiation ; and it is difficult to see how 

 the glass can be protected from radiation except by enclosing it 

 in another glass ; and under such circumstances its action will be 

 very feeble. 



Instrument-makers insist on the integrity of the mixture, the 

 composition of which they keep secret. This, of itself, ought to 

 exclude the instrument from the Meteorological Observatory. It 

 is possible that some storm-glass maker may reflect on the charac- 

 ter of my paper on the ground that the mixture adopted by me 

 is not the correct one. I have formed a number of these glasses 

 according to the recipe given, and find them to act very well. 

 The mixture must be heated, but gently, otherwise the sal-am^ 

 moniac will separate from the other ingredients. When poured 

 into the tube, corked and suspended, the glass may require a 

 day or two before it begins to act properly ; but a little shaking or 

 exposure to the sun, or thinning down by the addition of a small 

 portion of spirit or of water, will soon make the glass act properly. 

 Two tubes containing the same mixture, were placed, one in the 

 window, and the other in a test-glass within a foot of the win- 

 dow ; the first acted well, the second did not act at all, on account 



