226 PROGRESS OF METEOROLOGY IN 1889. 



Measurement of rain-fall. — Prof. Cleveland Abbe lias discussed at 

 length the seiirces of error in tlie collection and measurement of rain- 

 fall. From a study of published observations of rain-fall measures at 

 different heights above ground he finds that the deficit in collection 

 equals 6 per cent, multiplied by the square root of the altitude in meters. 

 It is further shown that the error of collection may be eliminated by 

 observing two gauges at different altitudes and applying a formula of 

 reduction. [Am. Meteor. Journal^ vi, p. 241.) 



Errors in thermometer readings. — Messrs. W. A. Eogers and R. S. 

 Woodward, in a paper on errors in reading mercurial thermometers, dis- 

 cussed an instrumental error hitherto generally overlooked. The theo- 

 retical study of Professor Woodward led to a dynamical equation ex- 

 pressing the motion of the end of the column in terms of its mass, the 

 volume and elasticity of the bulb, and the frictional and other forces, 

 including surface tension. When the latter forces are constant, and 

 when the rate of temperature rise is constant, the formula shows: 

 (1) That the motion of the end of the column is by pulsations of regular 

 recurrence, the reading of the thermometer being alternately too great 

 and too small ; (2) that the period of recurrence in the same thermom- 

 eter varies directly as the square root of the rate of temperature rise ; 

 (3) that the amplitude is sensibly constant; (4) that, since the ampli- 

 tude is nearly constant and the period thus dependent on the rate of 

 the temperature rise, the danger of error in thermometer readings is 

 greater for slow than for rapid temperature changes. {Report Am. 

 Assoc. A. 8., 1889.) 



Method of obtaining daily mean temperatures. — Dr. W. Koppen has 

 sought for the most accurate method of deriving the true daily mean 

 temperature from observations made at 8 A. M., 2 and 8 p. m., and the 

 minimum temperature. The author assumes, as recommended by the 

 Vienna Congress, that the true mean temi)eratnre m is given by the 

 formula m=n—lc {n — min) where « is the arithmetical mean of the three 

 observed values, and k a constant to be determined. The author has 

 determined the value of h for different seasons at different places where 

 hourly temperatures have been recorded, and finds that in spring and 

 summer the value is almost the same at all stations. In a similar way 

 the author investigates the accuracy of the combination of the 8 A. m., 

 2 p. M., and 8 p. M. without the minimum tem])erature. [Mctcorolo- 

 gische Zeitschrift, 1889, vi, p. fl].) 



Reduction of air pressure to high levels. — As a supplement to his paper 

 on the construction of isobars for the level of 2,500 meters {Meteo- 

 rologische Zeitschrift, December, 1888), Dr. Koppen gives the following 

 table of constants for obtaining the mean temperature of the air column, 

 or the temperature at 1,250 meters, by subtracting them from the sea- 

 level temperatures : 



