10 ROSCOE. — METEORIC DUST IN SNOW. 



coincidences need not be attended to, the arc should be carefully 

 observed five or six times distributed over the swing, or, which 

 would probably be found more accurate, the clock times should be 

 noted when the arc attains definite values, beginning with an arc 

 slightly greater than the greatest used for coincidences, and going 

 on till it is reduced to about one-third of its initial value. The 

 barometer and thermometer should be read at the same time. 

 The object of this is to determine the law of decrease of the arc, 

 and thereby render it possible, in the subsequent reduction of the 

 observations for time, to correct for the arc without assuming that 

 it decreases strictly in geometric progression. 



The geographical position, latitude especially, of the pendulum 

 station must be found, and the height above the sea level. The 

 geological character of the formation on which the pendulum 

 observatory is built should be stated. Should it be found impracti- 

 cable to erect the observatory on land, it may be built on the ice, 

 provided there be no sensible change of level of the ice, and 

 no motion of any kind, the alteration of which is not extremely 

 gradual, and provided also, that means can be employed for check- 

 ing the clock's rate by astronomical observations. Should the 

 pendulum be swung on ice, the depth of the sea at the place must 

 be measured. 



Twenty-four hours' observation with each pendulum would 

 give an excellent result, provided the weather permit of a trust- 

 worthy determination of the clock's rate. The days on which the 

 two pendulums are swung need not be consecutive. 



4. On the Detection of Meteoric (Cosmical) Dust in the 

 Snow of Arctic Kegions. By Prof. H. E. Roscoe, 

 F.RS. 



It has been shown by Nordenskiold* that pure snow collected in 

 the northern regions far distant from any source of dust, contains 

 small black particles left behind when clean snow is melted. These 

 black particles consist mainly of iron, but contain distinct quan- 

 tities of cobalt, thus proving their non-terrestrial character. It 

 would be very interesting to confirm these observations of the 

 wide-spread depositions of fine cosmical dust by a repetition of 

 the process adopted by Nordenskiold, which consisted in collecting 

 a large quantity of apparently pure snow, and allowing the same 

 to melt, placing it, for this purpose on a clean sheet, spread 

 out, and arranged so that the water should drain away, leaving 

 the black particles on the sheet. These should then be carefully 

 collected, when the greater part of the snow was melted, by placing 

 the remaining snow in a bottle or glass, and allowing it to melt 

 completely, when the black particles will sink to the bottom and 

 the clear water can be poured off. Or the black particles can 



* PoggendorflPs Annalen, 151, p. 154. 



