297-299] Classification 333 
The upper half of the Y-shaped diagram consists of two distinct branches, 
one of which is found to contain a far larger number of nebulae than the 
other. The principal branch contains the normal “ spiral ” nebulae, which are 
characterised by a circular nucleus from which emerge two (or occasionally 
more) arms of approximately spiral shape. Typical examples are shewn in the 
left-hand half of Plate X. These nebulae are subdivided into three classes, 
designated Sa, Sb, Sc, class Sa fitting almost continuously on to class E 7. 
The minor branch contains a special class of spirals characterised by the 
circumstance that the spiral arms appear to emerge from the two ends of a 
straight bar-shaped or spindle-shaped mass. These are sometimes called 
^-nebulae, on account of their characteristic shape, although 0-nebulae 
would appear to be a better designation. Typical examples are shewn on the 
right-hand half of Plate X. Both here and in Plate IX, both of which are 
taken from Hubbles original paper, the examples shewn are the actual nebulae 
which Hubble selects as the milestones to mark the various sub-classes. 
About 97 per cent, of known extra-galactic nebulae are found to fit into this 
Y-shaped classification. The remaining 3 per cent, are of irregular shape, 
and refuse to fit into the classification at all. These include the two Magellanic 
clouds and other star-clouds which Hubble treats as nebulae for purposes of 
classification. The irregular nebulae are distinguished by a complete absence 
of symmetry of figure and also by the absence of any central nucleus. Typical 
examples of “ irregular nebulae ” are shewn at the bottom of Plate IX. 
Apart from the irregular nebulae, Hubble states that, out of more than a 
thousand nebulae examined, less than a dozen refused to fit into the Y-shaped 
diagram at all, while in less than 10 per cent, of the cases was there any 
considerable doubt as to the proper position of a nebula in the diagram. 
Clearly then, the Y-shaped diagram provides a highly satisfactory working 
classification. 
Physical Interpretation. 
299. Obviously the proper physical interpretation of the classification just 
described is of the utmost importance to cosmogony. 
A first and most important clue is provided by the fact that numbers of the 
great nebulae are known to be in rotation. In 1914 Wolf* detected rotation 
in the spiral M 81 (see Plate XII, p. 351), and in the same year Slipher-f* 
discovered it in the type nebulae N.G.C. 4594 of class Sa (see Plate IX). 
Pease measured spectroscopically the velocity of rotation along the major axis 
of this latter nebulae in 1916 J and along the major axis of the Andromeda 
nebula M 31 in 1918 §. 
The symmetry of figure shewn by nebulae of the E and Sa types is precisely 
such as rotation might be expected to produce, and thus suggests an inquiry 
* Vierteijahrsschrift der Axtron. Gesell. lxix. (1914), p. 162. 
t Lowell, Obs. Bull. No. 60 (1914). + Proc. Nat. Acad. Sci. n. (1916), p. 517. 
§ Ibid. iv. (1918), p. 81.