354 SECTIONAL TRANSACTIONS.—A. 
nove, but two facts render the study of Nova Pictoris of particular interest : (1) The 
light changes have proceeded with abnormal slowness ; eleven weeks after discovery, 
the nova was brighter than when discovered and five years after discovery the 
magnitude had decreased only to 8:5, the pre-outbreak magnitude being 12™-75, 
(2) The bands are not so broad as they usually are in nove, being about 12A at 
i 4400; the emission band spectrum is thus not complicated by extensive over- 
lapping of bands and is comparatively simple to disentangle. 
A survey is given of the main spectral changes. A feature of interest is the 
appearance about October 1925 of forbidden emissions associated with metastable 
states of the ionised atom of iron. Such emissions are prominent in the spectra of 
1 Carine and of a few other stars. They all arise from transitions between even 
terms, ending at one of the two lowest energy levels. These emissions gradually 
fade away in the early part of 1926, at about the time that typical nebular and helium 
bands become strong. The N;, N, bands are late in appearing and never become a 
prominent feature in the spectrum. Up to 1926 March 13 no trace of ionised helium 
could be seen in the spectrum, but on March 15 the ) 4686 band was one of the strongest 
in the spectrum. Its intensity was very variable for several months after its initial 
appearance. The spectrum in 1928 was of interest on account of the development of 
a hitherto unobserved emission, of unknown origin, at about 6087, which became 
one of the strongest lines in the spectrum. At this stage, the neutral helium emissions 
were weak but the He II line at } 4686 was the strongest in the spectrum. In 1931 
the } 6087 emission had become the strongest line; Ha and an emission at ) 5724 
were also prominent. The higher members of the Balmer series had decreased con- 
siderably in intensity as compared with Hg. No trace could be found of the N,, No 
lines. The spectrum now approximates to that of a Wolf-Rayet star. 
The main phenomena are in accord with the hypothesis that the nova was in a 
state of rapid expansion up to the time of the first maximum and that at or about 
maximum, shells of gas were thrown off. By correlating the observed rate of 
expansion with the increase in brightness, a parallax of 0”:0015 is derived, corres- 
ponding to an absolute magnitude at maximum of —7:9. The radius of the nova at 
maximum was about 380 times the radius of the Sun. The progressive shift in the 
displacements of the two main absorptions for several weeks after maximum can be 
explained on the hypothesis that the absorptions are due to the absorption by 
expanding shells of gas of light emitted by the star. 
A summary is given of the observations at Johannesburg of the nova as a multiple 
star and of the surrounding nebulosity. The increase in the diameter of the latter 
is in good agreement with the hypothetical parallax of 0”-0015. Possible causes of 
the outbreak are considered ; the hypothesis that it is due to a sudden release of energy 
within the star, rather than to a collision with another star or to the star entering a 
nebulous region, seems to accord best with the various facts which have to be 
taken into account. 
Dr. 8. A. Mrrcnety.—The 1930 Helipse observed from Niuafoou. 
The October, 1930, eclipse was very spectacular in all of its details. The only 
accessible spot from which to observe it was a tiny volcanic island, called Niuafoou 
or Tin-Can Island, in the Tonga group, situated nearly on the 180th meridian and at 
15° south latitude. Only two scientific expeditions made observations, a small 
party from New Zealand and a better equipped party from the United States, the 
latter being under the auspices of the U.S. Naval Observatory. The weather, always 
uncertain in the tropics, behaved magnificently well at the crucial moment. ; 
This was my eighth eclipse. To have the opportunity of working during totality 
for the grand aggregate of fifteen minutes it has been necessary for me to travel no 
less than 90,000 miles. 
As always, my work was spectroscopic. At this eclipse I had two concave 
gratings, each without slit. History has shown that it is excessively difficult to 
secure eclipse spectra of perfect definition, especially when the spectrographs are 
used without slit. The reason for dispensing with the slit is to secure the heights 
to which the vapors extend in kilometers above the solar surface. At increased 
heights there is a slow diminution of temperatures and a more rapid fall in pressures 
so that conditions are readily reached where an atom can lose an external electron. 
When this happens, the atom is said to be ionised and its spectrum is radically 
different from that of the ordinary atom. Sir Norman Lockyer was the first to call 
