TOTAL ECLIPSE OP THE SUN, APRIL 16, 1893. 
571 
way of ensuring good definition from one end of the spectrum to the other is to get 
the spectrum of a star in focus, for collimation can only be effected over a small region, 
unless the collimating lens is perfectly achromatic. To make the image of the star to 
travel in a direction parallel of the edges of the prisms (in order to give breadth of the 
spectrum) it was necessary to throw the polar-axis slightly out of adjustment and to 
keep the clock going to star rate by means of the fine adjustment in right ascension ; 
after several trials a photograph was obtained which showed the spectrum of the star, 
Fig. 6. 
Dark slide employed in Brazil. 
but only between H /3 and Hy, and this seemed to be fairly well in focus. From this 
time up to the day of the eclipse there was no possibility of trying this again, so that 
the focus had to be taken as final. It so happened, however, that the spectrum was 
in focus between XX 58 and 44 and again between XX 37 and 35. The same thing 
appears to have occurred, only more seriously, when the instrument was previously 
used during the eclipse of 1883. 
Plates Employed. 
As stated in Mr. Fowler’s report, there was some uncertainty as to the kind of 
plate best suited for this work, and for that reason similar brands to those used by 
him were employed with this instrument, except that of the isochromatic ; those 
made by Edwards were solely used, the other makes proving (in the preliminary 
trials) not to be as suitable under the climatic conditions. About twelve of the 
4 D 2 
