138 



KNOWLEDGE, 



[July 1, 1892. 



manufacture of which is conducted with the most elaborate 

 precautions, not only against shock, but against the 

 smallest amoimt of friction. 



The protection of factories against lightning is a problem 

 of considerable ditficulty. According to Mr. Otto Gutt- 

 mann, whose recent paper before the Society of Chemical 

 Industry contains much useful information on this and 

 other matters connected with the dangers of explosives, a 

 system similar to that of Professor Lodge's "network"' 

 protector has been extensively and successfully used bj- 

 Austrian military authorities. The system is similar to 

 that by which electrometers are shielded from electriiication 

 by means of a wire cage, the building being covered by a 

 network of galvanized iron wire. This material is, of 

 course, much cheaper than copper, and its smaller electric 

 conductivity does not appear to be a serious drawback in 

 the case of electric discharges of such high potential as 

 that of lightning. 



RADIOMETRY. 



Bv A. Jameson. 



THE kinetic or, as it is sometimes called, the 

 molecular theory of matter, by which its sensible 

 qualities are referred to the motion of atomic and 

 molecular parts, and the undulatory theory of 

 light, which asserts that radiation is due to 

 transverse waves in a medium with which all space is 

 filled, have become thoroughly incorporated with modern 

 physical science. Employed m the first place as working 

 hypotheses, these theories have gradually become esta- 

 blished as truths. To give anything approaching a 

 complete exposition of them would be a most extensive 

 undertaking. But, without pretending to this, it is 

 thought that a discussion of some few typical instances in 

 which the operations of these laws have been recognised 

 may prove interesting. 



Prof. Crookes' radiometer, or light mill, furnishes a 

 remarkable example of the conversion of energy in the 

 form of ethereal waves into molecular and subsequently 

 molar motion. The commonest form of radiometer is 

 shown in Fig. 1. Four light vanes of mica, attached to 

 c radial wire arms, are fixed to a central 



cup, that is balanced, like that of a 

 compass needle, upon a fine steel point. 

 One side of each vane or paddle is painted 

 a dull black with lampblack ; and it is 

 upon these lampblacked faces of the vanes 

 that molecular pressure, resulting from 

 radiation, wiU be exerted. A small glass 

 tube is fixed verticaDy over the central 

 pivot cup. In the position shown this 

 tube is free from contact with the fly, 

 but serves, should the instrument be 

 inverted, to prevent it from toppling off 

 the needle-point. The glass bulb, having 

 been highly exhausted by a mercury 

 pump, is hermetically sealed. 

 Remembering that what we call heat is simply a state of 

 vibration of the molecular parts of bodies — restricted in 

 the case of solids and liquids, extending to perfectly free 

 excursions in the case of gases — we will proceed to consider 

 the effect of radiation, as, for instance, of difi'used daylight, 

 upon the piece of apparatus just described. Light traverses 

 the glass envelope freely, because, as we may assume, 

 there is no correspondence between the periods of the 

 luminous waves and those of the molecular vibrations in 

 glass. With the lampblacked surfaces of the vanes, 

 however, the case is vei-y different. Here the periods of 



Fis. 1. 



the vibrations accord so nearly with those of waves of 

 light that complete absorption takes place before an 

 appreciable thickness of the substance has been penetrated. 

 Just as the swing of a pendulum is amplified by properly 

 timed impulses, so the swing of a molecule responds to 

 transverse ether waves of suitable frequency ; and light 

 waves are quenched by the lampblack in consequence of 

 the conversion of their energy into molecular motion, that 

 is to say, heat. Mica, like glass, and like most gases, is a 

 transparent substance, whence it follows that light falling 

 upon the clear surfaces of the vanes will also be trans- 

 mitted to the layers of lampblack, and by them absorbed, 

 with corresponding elevation of their temperatures. So 

 little absorption takes place in mica that practically all 

 light falling on the bright sides of the vanes is either 

 transmitted as described or is reflected. But, in some 

 instruments, the mica surfaces are coated with a bright 

 metallic film, and in such cases, excepting only at the 

 blackened parts, nearly all of the incident light wiU be 

 reflected. Where light is not absorbed, of course it cannot 

 be the source of heat ; and therefore no increase of 

 temperature, corresponding to that which takes place 

 in the lampblack, can occur at the polished surfaces of 

 the vanes. Undoubtedly some heat will be conducted 

 through the mica from the blackened to the polished side ; 

 but since this material is an extremely bad conductor, the 

 amount of heat thus transmitted must be smaU. On the 

 other hand, heat will pass rapidly from the warmed 

 lampblack to the air or other gas with which it may be 

 in contact ; and it wUl presently be shown to be this kind 

 of transference of heat that accounts for the rotation of 

 the light mill. 



To make this matter clear we will suppose, m the first 

 place, that the glass bulb contains gas at the ordinary pres- 

 sure of the atmosphere, and we will confine our attention 

 to one only of the four similar and similarly situated mica 

 vanes. As the molecular vibrations at the surface of a 

 solid body, such as lampblack, will take place with most 

 freedom when normal to that surface, we may expect 

 every gas molecule that strikes the warmed face of the 

 vane to be thrown back in some direction, making a 

 smaller angle with the perpendicular than does the direc- 

 tion of incidence. Thus the lampblack will give rise to a 

 " molecular wind," made up of gas molecules whose paths 

 are, upon the average, nearly perpendicular to the heated 

 surface. But the mean free path of gaseous molecules at 

 the normal temperature and pressure is extremely smaU. 

 It is estimated by Maxwell, in the case of hydrogen, at 

 965 meter tenths (say lo-^o m^i-)' '"^ ^^^ ''^se of oxygen 

 and of carbonic acid at 560 and 430 meter tenths"' respec- 

 tively. Hence the molecules rebounding from the 

 heated surface wiU encounter others, and wUl thereby 

 have the direction of their motion altered many times 

 before they travel an appreciable distance ; and hence 

 the extra speed acquired by the gas molecules by collision 

 \\ith the hot lampblack — ^the extra temperatm-e, in other 

 words, of these free molecules — will be difi'used in every 

 direction through the atmosphere within the bidb. Now, 

 since the blackened surface of our vane is bombarded by 

 molecules having an average velocity corresponding to (say) 

 15° (or whatever may be the initial temperature of the 

 instrument) and since these molecules are thrown off again 

 with an average velocity corresponding to (say) 15-1°, it 

 miglit be thought that molecular pressure should be 

 developed, even in the circumstance we have supposed, 

 that is, in gas at the normal pressure of the atmosphere. 

 For action and reaction must be equal and opposite ; and 



* Im. tenth = 1 m. xlO-'" 



