THE ORIGIN OF THE EARTH. 21 



was organized on the collisional basis, nothing but a negative answer 

 seems possible. If the meteorites could be supposed to come together 

 so as to revolve in harmonious orbits about a common center, on the 

 planetary basis, the assemblage might be perpetuated, but this takes 

 the case out of the typical meteoritic class, and carries it over to the 

 planetesimal. 



Under the conditions of the case, it is not apparent how a meteoritic 

 nebula of the quasi-gaseous or collisional type can grow up de novo by 

 the assemblage of dispersed meteorites, or by the aggregation of chaotic 

 matter, if the material were endowed with the present momentum 

 of the system. 



2. Derivation from, Previous Aggregates. 



If the origin of the nebula be assigned to the dispersion of some 

 previous large body, three phases are conceivable: (a) dispersion by 

 explosion, (6) dispersion by collision, and (c) dispersion by tidal dis- 

 ruption. 



(a) Dispersion by explosion. — The diffuse state of the parent nebula 

 may perhaps be assigned to explosion, following the analogy of one 

 of the hypotheses of the origin of new stars, though it is difficult to 

 assign any probable and adequate cause for such explosion. But if 

 it takes place, the dispersing force must obviously be radial, in the 

 main, and after the matter has made its outward excursion and is 

 arrested by gravitation, it must return on nearly direct lines and col- 

 lide at the virtual point of departure. The result of this collision, 

 if the dispersion have nebular dimensions, must be the development 

 of enormous heat and the probable conversion of the whole into a 

 gaseous body. If so, the evolution must thence proceed along gaseous 

 lines. In this case, nothing proper' y analogous to a meteoritic con- 

 dition is likely to be developed in the system, except as a mere incident, 

 and the case does not really belong under the meteoritic hypothesis. 

 Such an explosion may indeed give rise to material shot away beyond 

 the control of the system, and this part might be truly meteoritic, but 

 it would be lost to the system, and fall under the case already discussed. 



(b) Dispersion by collision. — If the dispersion be assigned to the 

 collision of two large bodies, as postulated by Croll 1 and others, the 

 heat developed must be presumed to be great enough to convert the 



1 Croll, Climate and Cosmology, 1889, pp. 297-315. 



