334 
POPULAR SCIENCE REVIEW. 
remained a day or two in a dry atmosphere, it is placed over the magnet, or 
magnets, with its ends resting on slips of wood, so that the under surface of 
the plate just touches the magnet. Fine iron-filings, produced by a draw- 
filing” Norway iron which has been repeatedly annealed, are now sifted 
uniformly over the film of lac by means of a fine sieve. The spectrum is 
then produced on vibrating the plate, by letting fall vertically upon it, at 
different points, a light piece of copper wire. The plate is now cautiously 
lifted vertically off the magnet and placed on the end of a cylinder of paste- 
board, which serves as a support in bringing it quite close to the under sur- 
face of a cast-iron plate (1 foot in diameter, ^ inch thick), which has been 
heated over a large Bunsen-flame. Thus the shellac is uniformly heated, 
and the iron-filings, absorbing the radiation, sink into the softened film and 
are “ fixed.” He generally allows the heat to act until the metallic lustre 
of the filings has disappeared, by sinking into the shellac, and the film 
appears quite transparent. This degree of action is necessary when photo- 
graphic prints are to be made from the plate ; but when they are to be used 
as lantern- slides, he does not carry the heating so far. After the plate has 
cooled, it is allowed to fall upon its ends on a table, so that any filings 
which have not adhered may be removed. 
Seven Years' Magnetic Observations. — The Rev. J. J. Perry has published, 
in a paper lately read before the Royal Society, the result of seven years’ 
observations at Stonyhurst. He states that the yearly mean values of the 
horizontal force are found to vary progressively from 3-5926 to 3*6178 in 
British units, the mean for October 1, 1866, being 3*6034, with a secular 
acceleration of 0*0042. Calculating from the monthly tables the mean 
value of the horizontal force for the six months from April to September, 
and for the semi-annual period from October to March, we find the former to 
be 0*0005 in excess over the latter, showing, that this component of the 
intensity is greater during the summer than during the winter months. 
Treating the dip observations in a precisely similar way, we obtain 69° 45' 
21" as the mean value of this element for October 1, 1866, subject to a 
secular diminution of 1' 49" *2 ; the extreme yearly means being 69° 48' 47" 
and 69° 37' 52". The resulting excess of 10" for the winter months in the 
computed semi-annual means is so small that the observations tend mainly 
to show that the effect of the sun’s position is not clearly manifested by any 
decided variation in the dip. Deducing the intensity from the above 
elements, we obtain for the summer months the value 10-4136, whilst that 
for the winter months is 10-4128. The intensity of the earth’s magnetic 
force would thus appear to increase with the sun’s distance, but the differ- 
ence is not large enough to have more than a negative weight in the ques- 
tion under discussion, This weight, moreover, is lessened by the slight 
uncertainties arising from the probable disturbing causes at the first mag- 
etic station. 
Attraction caused by Vibrations of the Air. — In the “ Philosophical 
Magazine ” is a paper by Professor Challis, in which the author maintains 
that the condensation in waves propagated from a centre will vary inversely 
as the distance, and that the rate of diminution of the condensation or rare- 
faction with distance from the centre will be continually changed from the 
law of the inverse square of the distance to that of the simple inverse of the 
