Since then a great deal more observational evidence has accumulated, and an exhaustive investigation made by Dr Seares of Mount Wilson in 1922 leaves very little room for doubt that the motions of the stars shew a real, and fairly close, approximation to equipartition of energy. The table overleaf shews the final result of Seares’ discussion. The stars are first classified according to the different types of spectrum their light shews when analysed in a spectroscope. _Equipartition of Energy in Stellar Motions_ +----------------+-----------+-----------+-----------+-------------+ | | Average | Average | Average |Corresponding| | | weight | speed | energy | temperature | | Type of star | _M_ | _C_ | ½_MC_² | (degrees) | | | (grammes) |(cms./sec.)| (ergs) | | +----------------+-----------+-----------+-----------+-------------+ | Spectral | | | | | | type _B_ 3 |19·8 × 10³³|14·8 × 10⁵ |1·95 × 10⁴⁶|1·0 ×10⁶² | | ” _B_ 8·5|12·9 |15·8 |1·62 |0·8 | | ” _A_ 0 |12·1 |24·5 |3·63 |1·8 | | ” _A_ 2 |10·0 |27·2 |3·72 |1·8 | | ” _A_ 5 | 8·0 |29·9 |3·55 |1·7 | | ” _F_ 0 | 5·0 |35·9 |3·24 |1·6 | | ” _F_ 5 | 3·1 |47·9 |3·55 |1·7 | | ” _G_ 0 | 2·0 |64·6 |4·07 |2·0 | | ” _G_ 5 | 1·5 |77·6 |4·57 |2·2 | | ” _K_ 0 | 1·4 |79·4 |4·27 |2·1 | | ” _K_ 5 | 1·2 |74·1 |3·39 |1·7 | | ” _M_ 0 | 1·2 |77·6 |3·55 |1·7 | +----------------+-----------+-----------+-----------+-------------+ These different types of stars have very different average weights; the second column of the table shews that they exhibit a range of over 16 to 1. The third column, which gives the average speeds of these different types of stars, shews that the heaviest stars move the most slowly, and the lightest on the whole the most rapidly. The next column gives the average energy of motion of the different types of stars. This shews that the variation in speeds is just about that needed to make the average energies of all types of stars equal. An exception certainly occurs in the first two lines, which refer to the heaviest stars of all. Apart from these, the remaining ten lines shew a ratio of 10 to 1 in weight, whereas the average deviation of energy from the mean is only one of 9 per cent.