318 CARNEGIE INSTITUTION OF WASHINGTON. 



On this basis it would follow that, in proportion as any assemblage 

 of matter looses an excess of electrons and thus becomes dominantly 

 positive at center, it repels any similar positive assemblage. To the 

 extent of its value, this repulsion should offset an equivalent amount 

 of gravitative attraction, whatever the ultimate nature of that attrac- 

 tion may be. In the normal stellar state this, as remarked already, is 

 regarded as negligible, but it may plausibly be assumed that, in the 

 vast deployment of a star into a spiral nebula, there would be an un- 

 usual amount of electric dissociation, and further that the vastness of 

 the deployment would enormously increase the facilities for the escape 

 of electrons. In this case the remnant positive charges of the two dis- 

 persed bodies would offset a larger fraction of mutual gravitation 

 than if the stellar matter had remained concentrated as before in 

 spheres. To the degree that electric repellancy was thus developed 

 between the deployed nebulse, the restraint of the prodigious velocities 

 acquired in approaching periastron would be neutralized and cor- 

 respondingly higher velocities would be retained. 



4. Still another possible source of high nebular velocities may spring 

 from radiation pressure. So long as stars remain essentially spherical, 

 the radiation pressure which they exert upon one another is very small 

 compared with their mutual gravitation. If, however, stars are 

 explosively scattered into nebulae, it seems eminently probable (1) that 

 the radiation sent forth from them would be greatly increased during 

 and just after the explosive stage, and (2) that an increased proportion 

 of the radiation pressure would be caught by the companion nebulae, in 

 cases in which both stars are disrupted, and that the nebulae would be 

 driven apart in a higher degree than in the case of undisrupted stars. 

 Furthermore, stellar matter, in passing from the gaseous state in which 

 it was ejected into condensed aggregates following its enormous expan- 

 sion and its greatly enhanced radiation loss, would pass through 

 those dimensional stages in which radiation pressure is most effective. 

 In so far, therefore, as enhanced radiation pressure might offset gravi- 

 tation after the stars had passed periastron and become dispersed 

 thereby into nebulae, there would be an increment of velocity to be 

 added to the velocity previously acquired. 



5. In the cases in which nebulae are supposed to be produced by col- 

 lision, there is likely to be a conversion of energy of translation into 

 some form of dissociative energy with resulting loss of translatory 

 velocity. If this is effective, there should be a class of nebulae with low 

 velocities, and these should be associated with dynamic indices of col- 

 lision. This seems in a measure to be realized in the low velocities 

 of certain nebulae that bear forms which seem to imply a collisional 

 origin, such as the Great Nebula of Orion, the Fish Mouth Nebula, 

 and the Trifid Nebula. These form a class well differentiated from the 

 flat, symmetrical spirals, as also from the typical planetary nebulae 

 whose forms imply a less dispersive mode of genesis. 



