NATURAL PHILOSOPHY. 201 



decree of the depression of the mercurial column demands careful 

 attention, especially on the lakes, since the later or westerly gaL, the 

 rear of the cyclone, will be proportioned to this fall. When I he ba- 

 rometer has ceased to fall, the central portion is nearest; the local 

 change of wind will soon follow, and the barometer will commence 

 rising ; this is the most dangerous period of the storm, of which the 

 barometer has thus given us indication. All precautionary measures 

 should be taken, then, during the fall of the barometer. From 

 the favorable direction of the easterly or first winds of the cyclone, 

 navigators are tempted to sail up the lakes ; they should remember 

 that the extent and force of the later westerly winds may be esti- 

 mated by the fall of the barometer, but not at all by any moderation 

 of the first or easterly winds. The fall of the barometer on the lakes 

 is less than on the ocean from a like storm ; yet it is sometimes one 

 inch or even more, and whenever it exceeds the ordinary mean of 

 moderation storms, every precaution should be taken. 



ON THE GRADUATION OF STANDARD THERMOMETERS. 



The following account of the graduation of standard thermometers 

 at Kew Observatory, was given to the British Association, by Mr. J. 

 Welsh. In the year 1851, the Committee of the Kew Observatory, 

 impressed with the importance in meteorological investigations of pos- 

 sessing thermometers of a better class than those hitherto procurable 

 from opticians, took steps with the view of producing such instruments 

 under their own superintendence, for distribution to institutions and 

 individuals who might require accurate standards of reference. The 

 committee were furnished with the information necessary to carry out 

 their intentions, by M. Regnault, of Paris, who had been accustomed 

 to construct his own thermometers by a method proposed by himself, 

 and with an accuracy previously unknown ; they were also supplied 

 under his directions, with the requisite apparatus. It had been as- 

 sumed by physicists, that at all temperatures, as high at least as that of 

 boiling water, the apparent expansion of mercury in a glass envelope, 

 is uniform for equal increments of heat. A mercurial thermometer 

 might, therefore, be called a standard instrument, when the divisions 

 of its scale corresponded everywhere to equal volumes of the contained 

 fluid, and when the readings were known for the temperature of melt- 

 ing ice, and of water boiling under a certain barometric pressure. If 

 the tube were perfectly uniform in its bore, it would only be necessary 

 to make a scale of equal parts between the freezing and boiling points, 

 and to extend the division above and below these points, but as per- 

 fect tubes were in practice not procurable, it became necessary in 

 dividing the scale, to make allowance for the variations in the tube's 



C* ' 



capacity. These variations could be obtained by carefully measuring 

 a short column of mercury, (an inch or less in length,) as it is made to 

 pass along the tube by successive steps, each of which is as nearly as 

 possible its own length. In the thermometers constructed according 

 to M. Regnault's process, the divisions do not represent degrees of the 



