E. TERRESTRIAL PHYSICS AND METEOROLOGY 111 



centre of the revolving caps of the Robinson instrument, in- 

 stead of describing exactly one third of the path of the wind, 

 as they should do according to the theoretical researches of 

 that astronomer, differ from this to such an extent that the 

 indications of their dials must be multiplied by numbers 

 varying from 2.1 to 2.9 in order to deduce the true velocity 

 of the wind. The result most to be desired in relation to 

 this branch of anemometry is the determination of the rela- 

 tion between these varying factors and the actual dimensions 

 of the respective instruments. But Dr. Dohrandt finds, as 

 had Cavallero and Stowe before him, that no such relation 

 can be discovered ; only so much is easily seen, that the mag- 

 nitude of the factor is in more direct connection with the 

 diameter of the hemispherical caps than with the length of 

 the arms of the anemometer. In respect to the pressure of 

 the wind, Dohrandt experimented with six different speci- 

 mens of Wild's wind -pressure anemometer, which consists 

 simply of a very thin plane piece of iron or wood, or thin 

 cardboard, hanging by a very thin arm from a pivot, and 

 which is by the force of the wind pushed aside from its ver- 

 tical into an inclined position. The angular extent of the 

 deflection is measured in degrees by an appropriate divided 

 arc. In connection with this, he calls attention to the fact 

 that the indications of this, and of all pressure instruments, are 

 decidedly affected by the density of the atmosphere, as indi- 

 cated by the barometer. The comparisons of the indications 

 of this pressure gauge with the velocity deduced from the 

 standard anemometers seems to show that the simple instru- 

 ment of Dr. Wild suffices, by means of the table given by 

 Dohrandt, to determine the velocity of the wind within less 

 than about one half a meter per second. On account, there- 

 fore, of its simplicity and inexpensiveness, it is probable that 

 this instrument will be widely used throughout the world, 

 especially as it has now been introduced at all the Russian 

 meteorological stations, and at many of those in southern 

 Europe. WilcVs " B.epertorium" IV., art. 5. 



THE SELF-REGISTERING BAROMETER OF REDIER. 



Among the many contrivances brought forward during 

 the past few years for the self-registration of meteorological 

 instruments, that of Redier seems both economical and a})- 



