METEOROLOGY AXD ALLIED SUBJECTS. 257 



of the imperial German commission on standards of measure, an exhaust- 

 ive memoir on the comparison of mercurial thermometers. In his last 

 chapter he gives a collection of his results that may well form the basis 

 of a reformation in every meteorological service that has not already 

 attended to this important subject. He gives formulae for computing the 

 changes in zero point of a thermometer depending on time and on the 

 exposure to high or low temperatures. The requirements of a good ther- 

 mometer are given by him in great detail; among them we note an item 

 frequently overlooked, namely, the importance of selecting the proper 

 kind of glass ; it having been shown that thermometers of the same gla«s 

 behave similarly in respect to the reduction to absolute temperatures or 

 the air-thermometer, and also similarly in respect to the changes with 

 time and temperature. Especially happy would it be if there could be 

 introduced glass such as that used in the construction of one thermome- 

 ter investigated by Dr. Thiesen, which during the past fifty years has 

 experienced no sensible change in its fiducial points, and whose tempo- 

 rary variations with temperature are but one-fourth of those experienced 

 by other thermometers, and whose corrections to reduce to the air-ther- 

 mometer are remarkably small. The effect of changes in external pres- 

 sure (as in a liquid bath, or with a varying barometer) is appreciable and 

 is therefore not to be neglected. Dr. Thiesen for the first time separates 

 the error of graduation of the scale from the error due to the lack of uni- 

 formity of the tube as found by calibration. 



He adopts as the temperature of 100° C. that which is now coming to 

 be recognized by the meteorologists and physicists — namely, the tem 

 perature of saturate pure aqueous vapor under a pressure of a column of 

 mercury 760™" high at zero C, in latitude 45'=', and at the sea level; 

 thereby relinquishing the objectionable adoption by meteorologists of 

 Eegnault's laboratory as the normal locality. {Z. 0. G. M., XVI, 1881, 

 p. 290.) 



T. n. Stevenson has observed the effects on the velocity of the wind 

 of nearness to the earth's surface, by establishing an anemometer on a 

 staff fifty feet high. He finds the increase with altitude very regular 

 between 15 and 50 or more feet and represents the velocity by the or- 

 dinates of a parabolic curve, the abscissas of which are the heights 

 reckoned from an horizontal axis 72 feet below the surface of the earth. 

 It follows that all anemometers ought to be at a uniform height and not 

 less than 20 feet above the ground. 



From the known velocity [v.) at a known height {h.) we can approxi- 

 mately compute the velocity (Y.) at any other height (H.) by the formulie. 



Y = v 



h -f 72 



where, however, li. must be more than 15 feet and H not much more 

 tban 50 feet, {Z. 0. G. M., XVI, 1881, p. 310.) 

 Voller, in the proceedings of the Scientific Society of Hamburg- Al- 

 S. Mis. 109 17 



