TOTAL ECLIPSE OF THE SUH, APRIL 16, 1893. 
579 
less, because the lower part of the corona was eclipsed when these photographs 
were taken. 
Photographs about Micl-Eclipse, 
As in the case of the African series there are several of the photographs, 
Nos. 6 to 16, which show no chromosphere spectrum, although the spectrum of some 
of the larger prominences are present. Only the upper parts of such prominences 
are visible, and their spectra are very simple, consisting of hydrogen, helium, and 
the H and K radiations of calcium. 
The continuous spectrum of the corona in these photographs is very similar in 
appearance to that described in the earlier photographs, except that in those of 
long exposure it extends to the considerable height of 12' above the moon’s edge. 
By reason of the short focal length of the object glass used in Brazil the images of 
the coronal rings would be six and a quarter times brighter than those given with 
the African instrument, and for the same reason, together with the less dispersion, 
the continuous spectrum is five and a quarter times brighter. 
Consequently, as will be seen on reference to the Brazilian negatives taken in 
mid-eclipse (Plate 12), they are remarkable for the great intensity and definition of 
the 1474 K-ring, which from measurements extend to a height of 2' above the 
moon’s edge. An enlargement of the ring as it appears in the Negative No. 12, 
is given in fig. S, where it is placed alongside a reduced copy of a photograph taken 
Fig. 8. 
Comparison of the 1474 K-ring with the lower corona. , 
by Schaeberle in Chili. The latter has been selected because the exposure was 
relatively short enough to make the lower corona thus obtained comparable with 
the spectrum ring'at 1474 K. It will be seen that the prismatic camera has picked 
out the brightest parts of the corona, and where it is strongest the spectrum ring 
4 e 2 
