224 CARNEGIE INSTITUTION OF WASHINGTON. 



order of 0.1 y. in diameter. A cloud of such particles containing only 0.1 mg. 

 of matter in a column 1 cm. square and of a length equal to the thickness of 

 the cloud may be practically opaque, the brightness of stars seen through it 

 being diminished as much as 8 magnitudes. 



(2) If the dust particles are more than O.lju in diameter, the scattering will 

 be non-selective as regards wave-length. If their diameter is less than 0.05 n 

 the scattering will be strongly selective like that of a gas, but for equal quantities 

 of matter may be nearly 100,000,000 times as great. Stars seen through such 

 a cloud will appear red. Theoretically it would be possible with particles of 

 a given intermediate size to form a cloud through which a star would appear 

 blue, but the actual occurrence of these conditions is improbable. 



(3) Such fine dust will be repelled by radiation pressure from the sun and 

 from all stars except the fainter dwarfs of types K and M. In the case of the 

 brightest stars, especially those of type B, the repulsive force may be as 

 much as 1,000 times the gravitational attraction. The dust will be driven 

 out into the regions between the stars, and, since it can not be in true equi- 

 librium there, will ultimately escape to very great distances. 



(4) It is probable that this dust forms the principal absorbing material in 

 the "dark nebulae" which obscure portions of the Milky Way. Such clouds 

 probably are held together by gravitational attraction; if so, their masses 

 must be considerable, of the order of 100 times the sun's mass for a small 

 cloud 1 parsec in diameter. In such a mass most of the material might be 

 in other states of aggregation, but the collisions which would occur within 

 it would supply the fine dust required to make it opaque. 



(5) Extended luminous galactic nebulae appear to owe their brightness to 

 illumination of dark nebulae by stars, as Hubble's investigations have shown. 

 In some cases the starlight appears to be reflected by opaque matter, such as 

 fine dust; in others it seems to excite monochromatic gaseous radiations, 

 probably by some process analogous to fluorescence. In either case it is 

 probable that the observed luminosity is superficial and confined to the star- 

 lit surface. According to this view, as wisps of dark nebulosity drift into the 

 illuminated region near a suitable star they appear as luminous nebulae. The 

 motions known to exist in the Orion nebula suggest that its appearance will 

 be entirely changed by such drifting in less than a million years. The nebula 

 doubtless will persist, but it will be formed of new wisps and clouds which 

 have drifted into the field of illumination of the stars dominant in this region. 



Internal Motions in Spiral Nebulae. 



Mr. van Maanen has completed measurements of three more spiral nebulae, 

 N. G. C. 2403, 4736, and 5055. These show results similar to those found 

 previously for M 33, 51, 81, and 101, and point more toward a motion outward 

 along the arms of the spirals than toward a rotational motion. If such is the 

 case, it will not be possible at present to use these values for a derivation of 

 the masses of the spirals. 



Several methods of discussing the material available have been used by 

 Mr. van Maanen in an attempt to gain some knowledge of the probable dis- 

 tances of the larger spiral nebulae. The following probably deserve the greatest 

 confidence : 



(1) Jeans's method, based on the conditions necessary for the breaking up 

 of the line-elements of filaments thrown off by the nucleus. 



