comets'' tails, the cokona, and the aurora. 181 



But at the surface of the sun the pressure would mount up to '2.75 

 niilligranis per square ceutinietpr. On the other hand, a cul)ic centi- 

 lueter of water, which vveiohs 1 o-ram at the surface of the earth, 

 would weig'li 27.47 o-rjuus at the surface of the sun — i. e.. the attrac- 

 tion of the sun would draw it inward with about 10,0()() times the force 

 with which the sun's light would tend to drive it away. 



Ver}' difterent is the cas(^ if. instead of a cul)ic centimeter, we con- 

 sider a much smaller cube. The pressure on its base would fall ofi' as 

 the square of its edij-e. but the weio-ht would diminish as tiie cube. 

 There must come a point at which the ])r(\ssure of the liyht would just 

 balance the weight; and still smaller particles would be driven off with 

 a force greater than their weight. Thev would behave, in fact, as if 

 gravity had become negative. 



For example, a cube of water measuring one-thousandth of a milli- 

 meter (10~^ cm.) in the edge would weigh 27.47 X 1(>~^'- gnis. ; and the 

 pressure of light on its base would be 2.75X 1()--^X !«»-% — 27.5xl()-^- 

 gms. — i. e., slightly more than its weight. 



In measuring wave lengths of light physicists denote one-thousandth 

 of a millimeter by the symbol /'. The critical value of the vdge of a 

 cube of water, i. e., the value for which its weight is exactly neutralized 

 b}^ the pressure of light at the sun's surface, is thus approximately //. 

 For a spherical drop the critical diameter may be calculated to be 1.5 // 

 for water. For other sul)stances the critical value is inversely propor- 

 tional to the specific gravity. 



A similar effect of extreme minuteness is familiar to us as the expla- 

 nation of the long time required by very small particles to settle 

 through the atmosphere, amounting to many months in the case of 

 the tinely-divided dust thrown u[) during- the eruptions of Krakatoa; 

 but the resistance to suspended dust particles can never exceed their 

 weight, since it is only called forth by the motion produced by the 

 weight itself. The pressure of light now" considered may enormously 

 exceed the weight, provided the paiiicles are small enough. 



Fi'om the motions, and especially the curvature, of comets' tails the 

 magnitude of the repulsive forces to which they are subject may be 

 calculated. Thus Bredichin tinds in fo'ur instances that the repulsion 

 must have been about 18.5, 3.2, 2, and 1.5 times the sun's gravitational 

 attraction. Now, the vapors emitted by comets are largely hydro- 

 carbons of specilic gravity al)out 0.8. To account for these repulsions 

 on Arrhenius's principle the drops nmst have had diameters of 0.1 /<, 

 0.59 ju, 0.94 yw, and 1.25 //, respectively. In another case, where the 

 tail curved toward the sun, Bredichin foimd the repulsion to be 0.;') 

 times gravity. This would indicate partitdes of diameter (> //. Par- 

 ticles of this ordei' of magnitude, and far smaller, arc familiai' enough 

 to us, especially in combustion and in the early stages of condensation. 



The theory suggested is, then, as follows: As the comet approaches 

 the sun the intense heat causes a \ ioleiit eru])tioii of hydi'ocai'bon 



