350 SECTIONAL TRANSACTIONS.—A. 
Bullen include fifteen in their tables. Many of these were first identified 
by Gutenberg, who realised that waves striking the boundary of the earth’s 
liquid core would be in part reflected and in part transferred from waves of 
compression to shearing waves or vice versa. 
The great majority of earthquakes have their foci at depths less than 
so km., but there are others which have very deep foci. Turner insisted 
that this was the only possible interpretation of some of the observations 
discussed in the I1.S.S. In the last five years much attention has been 
devoted to deep-focus earthquakes. Scrase and Stoneley demonstrated 
that such earthquakes can be recognised on our seismograms. Wadati and 
other workers in Japan find that the deep-focus earthquakes are distributed 
in a very remarkable way, indicating that there are well-defined flaws in the 
earth at depths of a few hundred kilometres, far below the basic level of 
isostatic compensation. 
This is the most spectacular of the numerous recent advances in 
seismology. 
DiscussION on the above paper. 
Dr. A. T. J. Dotzar ; Prof. G. R. Gotpssroucn, F.R.S.; Dr. H. 
JerFREYS, F.R.S.; Mr. Cosmo Jouns ; Mr. E. 'TILLOTSON. 
Tuesday, September 10. 
Discussion on New stars (10.0). 
Dr. H. SPENCER Jones, F.R.S.—General phenomena of new stars. 
Definition of a nova. 
Historical nove and bright nove of recent years. 
General sequence of events in a nova outburst ; the light curve and 
spectral changes. 
The absolute magnitude at maximum brightness. 
Evidence for expansion of outer atmosphere of star and for ejection of 
one or more gaseous shells. 
The later stages of a nova outburst. 
Nove in the Andromeda nebula. 
Frequency of nova outburst and possibility that every star passes through 
the nova stage. 
The explanation, by Milne’s theory, of the nature of a nova outburst. 
Mr. J. P. M. Prentice.—The discovery of Nova Herculis, 1934 (10.20). 
Prof. F. J. M. Stratron, D.S.0., O.B.E.—The spectral changes of 
new stars (10.40). 
Before the principal maximum the usual spectrum is of « Cygni (or 
earlier) type, the absorption lines being markedly displaced to the violet 
and bordered on the red side by weak undisplaced emission lines. As 
maximum brightness is approached the displacements diminish and the 
spectrum becomes of later (or less highly ionised) type. After maximum 
the reverse takes place: displacements increase to the violet, multiple 
spectra are formed with different displacements and of different types— 
earlier types and larger displacements coming together. The absorption 
spectra give way gradually to emission spectra as the star becomes fainter ; 
the bright bands are usually complex in structure, the radial velocities of 
