August 24, 1SS3.] 



SCIENCE. 



•2\') 



c<l. or is iinix?ll«Ml witli great enei-gy fi-om the 

 terminal of :ui imliictioii-coil in :i (icislcr's 

 tiilic. it is not ncfcssiirv to assuini.' thai tiic 

 niolwules ai-e made to vibrate in wliolly new 

 peritxls, hilt tliat tiie ainplitndc of thoir vibra- 

 tions in any and all periods lias been in- 

 creased. thereby giving greater an)plitude, and 

 consequent enei^y, to the radiant uudulatiuus 

 emitted. suHieient to atteet the eye. 



AVhen one oonsidi'rs the kinetic energy of 

 molecules due to their lemi)erature, it seems 

 probable that all l)odies — solid and liquid, as 

 well as gaseous — must be vibrating in all pos- 

 sible peritnls eontinuously ; but in solid sand iu 

 liquids the shortness of the free paths makes 

 interference too frequent to allow any mole- 

 cule to vibrate many times between imp.iets, 

 and henc-e the harinouies suffer most, and are 

 destroyed before they can have given rise to 

 undulations in sufficient number or in ami)li- 

 tudc to iicrform any optical service. By. heat- 

 ing a solid, greater amplitude is given to all 

 the vibrations, and we see the red or longer 

 undulations tirst during the process of heating, 

 because such are less easily destroyed by im- 

 pact than the shorter ones, which cannot have 

 at best so great an amiilitude. This state- 

 ment assumes that it is with molecules as it is 

 with visible masses of matter: the greater the 

 number of vibralions possible to it, the less 

 the possible amplitude. 



With these conditions as stated, it is readily 

 seen why common objects are not at all times 

 visilile. th.it is to say, are not hnninous. It is 

 because our ej-es are not sensitive enough to 

 respond to the slight energy of the undula- 

 tions due to both lack of amplitude and short- 

 ness of the rays, not because those rays are 

 absolutclv wanting. A. E. Dolbear. 



RADIOMETERS WITH CURVED VANES. 



Amono the radiometers in a collection which 

 I have recently examined were two with curved 

 vanes of silver. The mdius of curvature was 

 less than 2 cm. When placed in front of a 

 lamp, the concave side moves towards the 

 source of heat. I have found no satisfactory 

 explanation of these movements. According 

 to a recent article by Dr. Pringshoim, the 

 convex side of these vanes is sujjposed to be 

 at a higher temperature than the concave side. 

 The grounds for such an hypothesis are not 

 obvious ; and it would seem hardly possible 

 that an apprecialile ditl'crence cojild exist be- 

 tween the surfaces of a thin sheet of silver. 



It is more probable that the air on one side 

 of the vane is hotter than that on the other. 



Since the ' kick ' of a nmlecule tle|iends on 

 its increase in temperature, the vane will 

 move towards the side on which the air is the 

 warmer. 



Dr. Pringsheini mentions an experiment in 

 which he brought the lieat to a focus inside 

 the radiometer at a point in front of the vane, 

 lie found that the air gave no evidence of 

 being heated. I repeated the experiment 

 with solar heat, using a lens of three inches 

 diameter and four inches focal length. The 

 heat in air was sufficient to ignite instantly a 

 common parlor match. When the focus was 

 kept in front of the vane of an ordinary radi- 

 ometer for two minutes, no appreciable effect 

 was observed : the instant it touched the vanes, 

 however, the\- gave a start, anil began to re- 

 volve. This experiment shows that the effects 

 observed with curved vanes cannot be attributed 

 to concentration of heat-rays from the vanes. 



According to the kinetic theory, this rota- 

 tion is set up only if the molecules arriving 

 on the convex side of the vane receive a greater 

 positive increment to their velocity than those 

 arriving on the concave side. These conditions 

 are satisfied in this way : if the vanes are 

 wanner than the air, the particles leaving the 

 vane in both directions have an increasetl velo- 

 city ; but take, for inst.nncc, the particles 

 moving in lines parallel to the axis of the 

 concavity towards the vane from either side, 

 those on the convex side are scattered by re- 

 flection, those on the concave side are brought 

 to a focus at a distance (in this instrument) 

 of less than 1 cm. from the vertex of the 

 concavity. The molecules in the vicinity 

 of this focus receive an increase of kinetic 

 energy ; and similar rea.soning holds for the 

 sets of molecules moving parallel to each other 

 iu any other direction. Hence the molecules 

 on the concave side are hotter than those on 

 the convex side, though not necessarily so hot 

 as the vane itself. Since the molecules on 

 the concave side receive a smaller increase of 

 velocitj- from the vane, thej- give it a smaller 

 reactive push. ^ 



The action of the case in a radiometer is 

 vor3- prettily shown by wetting it with cold 

 water. The action is best examined with 

 curved vanes, or with vanes of metal covered 

 on one side with mica. The rotation is at 

 first in the same direction as on heating, show- 

 ing that the air has become cooled by contact 

 with the glass, but is alter a time reversed, 

 showing, that, by quafti-conduction through 

 the air, the vanes have become cool, while the 

 glass is regaining its original temperature. 



Geokue W. Evaks. 



