234 ANNUAL, REPOET SMITHSONIAN INSTITUTION, 1931 



We have so far spoken of the nebular equator as being of circular 

 shape, as it undoubtedly would be if the nebula were alone by itself 

 in space. But an actual bursting flywheel, of course, first breaks at 

 its weakest point ; if it were of absolutely uniform strength it would 

 begin to break at all points of its circumference at once. In the same 

 way, if the equator of the nebula were a perfect circle, and if the 

 substance of the nebula were disposed symmetrically around its axis 

 of rotation, the ejection of matter would necessarily start from all 

 points of the equator simultaneously; there could be no conceivable 

 reason why it should start at one point rather than any other. 



In nature we do not expect to find perfect balances of this kind ; if 

 the main factors are of exactly equal weight, some quite minor fac- 

 tor invariably intervenes to turn the balance in one direction or an- 

 other. In the present problem there could be no choice as between 

 one point of the equator and another if the various minor factors 

 were absent, but as soon as minor factors come into play, a discrimi- 

 nation at once takes place. 



We have so far spoken of the rotating nebula as though it were 

 alone in space. Yet it must have neighbors, and these will raise tides 

 Qii its surface, just as the sun and moon raise tides on the surface of 

 the rotating earth. Wherever the neighbors are, there will always be 

 two points of high tide antipodally opposite to one another and two 

 points of low tide intermediate between the two points of high tide. 

 The equator will not be strictly circular, but slightly elliptical. 



It is in all probability this tidal pull that determines the choice of 

 points for the ejection of matter. Matter will be ejected at the points 

 at which the gravitational pull of the nebula is weakest, and so at 

 the two ends of the longest diameter in the equator of the nebula. 

 After the nebula has passed its critical landmark, it ought still to 

 retain the lenticular figure which formed the landmark, but with the 

 additional feature of matter streaming out from two antipodal points 

 on its equator. 



This is exactly what we see in the types of nebulae which we de- 

 scribe as " spiral." In N. G. C. 5866 we see a nebula in which the 

 ejection of matter is probably just beginning; we notice the bulge 

 along the equator and a dark band which probably represents ejected 

 matter which is alreadj^ cooling. A more advanced state of develop- 

 ment is shown in N. G. C. 4594; and a still later one in N. G. C. 891 

 in which the ejected matter already dwarfs the central nucleus in 

 size, although probably not in total mass. 



These are all photographs of nebulae seen very approximately 

 edge-on. The well-known " whirlpool " in Canes Venatici (M. 51) is 

 a spiral nebula which may be very similar physically to that shown 

 in N. G. C. 891, but is seen face-on ; we are looking along its axis of 

 rotation. Again, the central nucleus occupies only a small part of 



