1887.] On the- Total Solar Eclipse of August 29, 1886. 181 



graphs obtained in India during the eclipse of 1871 show about the 

 same fineness of detail as those obtained in 1882 and 1886, but the 

 aperture then employed was only 2 inches. It appears from this 

 either that the corona does not as a rule show sufficient detail to 

 warrant an aperture greater than 2 inches, or that for some other 

 reason the full advantage of an aperture of 4 inches has not been 

 realised. 



A linear object having an angular magnitude of 6 seconds of arc in 

 a camera of 63 inches focal length would have a length of about 

 the 600th part of an inch in the photographic plate. Remembering 

 the fact that diffraction gratings have been photographed with 17,000 

 lines to the inch, it is clear that it cannot be the fault of the photo- 

 graphic film if greater definition is not obtained. 



The different adjustments which it is in the power of the observer 

 to regulate can all be easily made to the required accuracy. I 

 feel certain, at any rate, that during the last eclipse the orienta- 

 tion of the instrument and the average rate at which the clock went 

 during the eclipse were sufficient to give the theoretical resolving 

 power. There seems to me to be only one way in which an improve- 

 ment in the photographs can be hoped for. The mechanical arrange- 

 ment by means of which the clock motion is transferred to the camera 

 is far from perfect. The series of cogged wheels which transmit the 

 motion must do so, in my opinion, in an irregular and oscillatory 

 manner, so that although the average rate of the clock can be regu- 

 lated with sufficient accuracy, there are minor periods which seriously 

 interfere with the regularity of the motion and the definition of the 

 images. If we are to obtain better photographs of the corona we can 

 only hope to do so by means of a better mechanical arrangement for 

 moving the camera. 



Photographs of the Spectrum of the corona were obtained by means 

 of two instruments, one being identical with that employed at 

 Caroline Island in 1883. This spectroscope has two prisms of 62° 

 refracting angle, the theoretical resolving power being about 10 in 

 the yellow. (The unit of resolving power is that resolving power 

 which allows of the separation of two lines differing by the thou- 

 sandth part of their own wave-length.) The slit of this spectroscope 

 was placed so that it was tangential to the image of the sun formed 

 by the condensing lens. One plate was exposed during the whole 

 of totality. The results are good ; a number of lines belonging to the 

 prominences and to the corona are very distinct and can be measured 

 with fair accuracy. The difficulty of measurement lies in the multi- 

 tude of lines. I have measured at present between forty and fifty 

 distinct corona lines between the solar lines F and H, and a number 

 remain unmeasured. 



A comparison between the photographs of 1882 and 1886 shows 



