AucxUST 28, 1891.] 



SCIENCE. 



121 



Suppose a stream of ether of such material should sweep 

 radially sunward, its particles colliding in uuelastic impact 

 with the earth, what velocity must be given to this current 

 in order that the earth might be kept in its present orbit? 

 The velocity turns out to be eight millions of times the ve- 

 locity of light. The mass of ether colliding per second 

 would be 14,000 tons, which is equal to the mass of a sphere 

 i>f water having a radius of about fifty feet. 



But in LeSage's hypothesis the ether particles do not move 

 in parallel stream-lines. They plunge into the earth on all 

 sides, the sheltering efSect of each gravitating body upon the 

 other being the cause of gravitation. But this sheltering 

 effect is very small, by reason of the open structure that 

 matter must be assumed to have in order that the interior 

 particles of large masses may be accessible 'to direct impact. 

 It follows that the percentage of particles really effective in 

 producing gravitation must be very small ; likewise that the 

 individual particle velocities must enormously exceed the 

 velocity computed for a stream of ether sweeping radially. 

 :sunward and capable of holding the earth in its orbit. 



It is unphilosophical to condemn the theory of LeSage 

 because it requires us to deal with such immense velocities. 

 Any theory of gravitation must involve something unusual, 

 and it was pointed out by Laplace that the velocity of gravi- 

 tation must enormously exceed that of light. But there are 

 other difficulties. 



The rebound of the particles must be a perfectly elastic 

 collision; otherwise the bombarded body will rise in tem- 

 perature. By reason of the open structure assumed and 

 necessary, in order that the effective surface may be propor- 

 tional to mass, the exterior figure is of no importance. For 

 simplicity of conception, if we assume a solitary sphere in 

 ispace, it will be symmetrically beaten from all sides. The 

 particles which pass straight through, without deflection or 

 •elastic rebound, will be symmetrical all around, as will like- 

 wise the few which suffer reflection. A second body now 

 appearing in its field will shield the first by deflecting parti- 

 cles which would otherwise strike it, but will reflect to the 

 body an equal number which would otherwise not strike it, 

 the latter group having the same average momentum as the 

 former. 



Sir William Thornson has suggested that the difficulty 

 may be avoided by assuming that the collision is not a per- 

 fectly elastic one, and that the rise in temperature may be 

 prevented by rotation of the colliding particles. This might 

 be true of the ether particles, but there is the best of reason 

 for believing that there is no molecular rotation in solid 

 bodies. 



It seems probable, therefore, that a rebound sufficiently 

 xiaelastic to account for gravitation must result in a rise of 

 temperature, which can scarcely be admitted. The theory 

 has, however, been defended with great skill by Preston, 

 who has attempted to show that such a medium may even 

 produce the transverse vibrations of light. The objection 

 that gravitation must travel at enormously greater speed than 

 light he tries to meet by the hypothesis that the gaseous ether 

 may have two groups of particles, one much larger than the 

 other. The properties of the luminiferous ether have, how- 

 ever, been so well worked out in the last four or five years, 

 that it seems hardly probable that the gaseous ether can be 

 admitted as the medium which transmits light. Whether 

 gravitation can or cannot be explained in some such way as 

 LeSage suggested, it seems worthy of question whether the 

 gravitation medium can be ether which transmits radiant 

 tenergy with a velocity of 300,000,000 of metres per second. 



Sir William Thomson's last word on the elastic solid theory 

 of ether, according to which the compression wave is impos- 

 sible by reason of a property which is imparted to the me- 

 dium, seems to cut off the last hope that the elastic solid 

 luminiferous ether can be concerned in gravitation. The 

 electric theory of light does not require Sir William Thom- 

 son's limitation to be put upon the medium. According to 

 this theory the medium may be incompressible, and there is 

 strong reason to believe that it is practically so. As Willard 

 Gibbs has pointed out, the two theories seem to be practi- 

 cally on the same footing, if the third wave is given an infi- 

 nite velocity in the electric theory, and zero velocity in the 

 elastic solid theory. 



There are other points concerning the action of matter 

 upon the ether which are perhaps in a fair way to receive a 

 clearer solution. The observed fact that light travels in 

 water with a speed of about three-fourths of what it has in 

 air, apparently means that the transmitting medium is either 

 more dense or less rigid in water than in air. Fresnel's hy- 

 pothesis is that its rigidity is the same in the two media. 

 His formula, as developed by Eisenlohr, for the relative mo- 

 tion of ether and matter which it permeates, when the matter 

 is set into motion, assumes, clearly and baldly, that the ether 

 is more dense inside of matter than in free space. The 

 amount of ether occupying a volume of one cubic centimetre 

 will condense to nine-sixteenths of a centimetre on passing 

 into wal3r. It is compressed until its density is nearly dou- 

 ble. To be more accurate, its density increases by seven- 

 ninths of itself upon passing into water. Of course this is 

 to be regarded as a mathematical Action serving to bridge 

 over a gap in our knowledge of the physics of the ether. 

 Certainly a medium behaving in this manner would not be 

 considered to be a shining success as an incompressible me- 

 dium. Fresnel's conclusion rests mainly on an experiment 

 first made by him and repeated with great success in an im- 

 proved form by Michelson and Morley. This experiment 

 was to determine the effect of moving water upon the ve- 

 locity of light transmitted along its stream-lines. The re- 

 sult reached was that the resultant velocity of light is its 

 velocity in the quiescent liquid plus or minus seven-six- 

 teenths of the velocity of the moving liquid. 



The velocity of the water current was varied between 8.72 

 metres and 5.67 metres per second, in Michelson and Morley's 

 ex])eriment. A series was made with an intermediate velocity 

 of 7.65 metres per second. The weights of the three deter- 

 minations are quite different, and it appears to be still an 

 open question whether the result obtained is independent of 

 the velocity of the water. In Eisenlohr's analysis he assumes 

 a prism of matter moving bodily through a mass of quiescent 

 ether. In Michelson and Morley's experiment the water 

 was fed from an upper to a lower tank, passing on its way 

 through the experimental tubes. The conditions of the two 

 experiments do not seem to be necessarily the same. The 

 bounding surface between air and water is moving with a 

 very small velocity in the apparatus of Michelson and Morley, 

 and the observations are made through a fixed region of 

 space. It is not clear that this difference is of importance, 

 but it seems possible that it may be, in determining the effects 

 for different velocities. 



The close agreement between the observed value of the 

 velocity coefficient for the moving ether and the value com- 

 puted on the assumption of an actual condensation of the 

 ether, is, of course, a very worthy consideration. Still it 

 seems very improbable that such a condensation can really 

 take place. The ether may lag behind the moving water 



