December 28, 1883.) 



SCIENCE. 



819 



can be easily put in rotation ; i is a needle to rest in 

 a hole in a piece of lead, to prevent oscillation. The 

 dimensions given are: a b. 70 ram. ; 6 c, 3l) mm. ; d/. 

 1!' mra. The tubing openings were 8 mm. long and 

 6 mm. in diiimeter. 



The sound-radiometer (fig. 4) is readily made. 

 In cardboard about 8 mm. thick, holes are punched 

 at intervals of 6 mra. with the punch of the form 

 shown at A. When prepared in this way, the card- 

 board will be repelled if presented to the resonator 

 with the small ends of the holes toward it, and at- 

 tracted when reversed. To make these effects more 

 marked, the punch and die shown at B and C may be 



used on moistened cardboard to form 

 conical holes with cylindrical ends. 

 The conical holes alone show no 

 effects. By arranging the pieces of 

 pasteboard as in D, or better as in 

 E, a rapid rotation may be obtained. 

 The apparatus shown in fig. i is Fir.. 4 



called a sound-wind-mill. A Helm- 

 holtz resonator, a b, is placed before the opening 

 of the bo.^-resonator. Out of the smaller end, n, 

 a stream of air will be blown when the fork is vi- 

 brated, and its existence shown by the rotation of 

 the windmill, h t. The dimensions of the Helm- 

 holtz resonator for G are : diameter, 80 mm. ; the 

 opening at b, IG mm. ; at «, 2 mm. This last is very 

 important. It seems odd that the resonator with 

 two openings may be repKaced by such as shown at 

 R with only one. The opening may f.ice in any di- 

 rection, provided the windmiii is suitably placed, and 

 still the mill will turn. When the opening is turned 

 toward the resonator-box, tlie distance between the 



resonators may be as great as half a metre. The di- 

 mensions and form of the ball are important. A suit- 

 able one may be made by grinding off the top of a glass 

 globe 50 mm. in diameter, and covering the opening 

 with a very thin metal plate in which there is a hole 



3.6 mra. across. The puffs of air coming from the 

 opening I are vortex-rings, which may easily be 

 shown by filling the ball with smoke. 



By putting one of the wings of the sound-radiome- 

 ter before the box-resonator with the larger ends of 

 the holes facing it and at a distance of 2 cm. from it, 

 the mill may be made to rotate by the puffs of air com- 

 ing through the holes, which should be numerous. 



AURORAL EXPERIMENTS IN LAPLAND. 

 Mk.J. Rand C apron gives a brief account {The 

 obsermloru, Sept.) of Professor Lemstnim's experi- 

 ments, quite similar to that 

 which has already appeared in 

 f, SciKNCE. He thinks that 



Professor Lemstriim's conclu- 

 sion, that the height of auro- 

 ras " has been generally over- 

 estimated may probably open 

 a lively discussion, as un- 

 doubtedly his dictum will 

 that 'measurements of an 

 aurora on a long base must 

 be erroneous, as the observers 

 never see the same aurora.' " He thinks, 

 too, that the relation Professor Lemstrom 

 believes himself to have proved, between 

 movements of atmospheric electricity and the 

 •variations of the m.agnetic elements," may 

 be only apparent. 

 Mr. Capron believes that the experiments 

 described did " collect and m.ike apparent to the eye a 

 true auroral glow, its spectroscopic character being at 

 the same time tested and defined by experienced ob- 

 servers." He adds, " Yet one cannot help feeling 

 something of regret that, if only for further assur- 

 ance, the wave-length of some one line seen was not 

 (as far as we are aware) absolutely determined, on 

 some occasions at least, and that the observations 

 appear to rest only on a small instrument presumably 

 without scale." 



Mr. Capron's article is Important mainly for call- 

 ing renewed attention to the phosphorescence, or 

 fluorescence, theory of the principal (yellow-green) 

 line of the aurora spectrum. This theory, first 

 proposed by Angstrom, was advocated in the Phi- 

 losophical magazine for April, 1875, by Mr. Capron, 

 who is inclined to attribute the line to phospho- 

 rescence, apparently on the following grounds: 1°. 

 The ' phosphorei'cent appearance ' of the aurora ; 

 2°. The fact that phosphoiescencc is capable of giv- 

 ing quite sharply defined spectral lines^ as shown 

 by his observations with a phosphorescent vac- 

 uum tube ; 3°. The fact that the auroral line 

 belongs to 'the principal region of phosphores- 

 cent light;' 4°. 'The observed circumstance that 

 the electric discharge has a phosphorescent after- 

 glow.' 



Mr. Capron observed, moreover, that the auroral 

 line lies in the region of a certain bright band in the 

 spectrum of a phosphoretted hydrogen fiaine, though 

 somewhat nearer the red end of the spectrum than is 



