534 0. PIAZZI SMYTH ON THE 



tinuously as well as speedily obtained at one temperature, may have been a 

 slowly accumulated piece of patchwork at several temperatures. 



Other portions again have been suggested to be owing to the natural dis- 

 placement of lines by the rotation of the Sun, when taken from either East or 

 West limb, instead of from the centre of the visible disc. But these effects 

 could only have been exceedingly small ; for I had much difficulty in realising 

 anything of that kind to be measurable, when I actually tried the experiment 

 with the apparatus arranged as employed through the whole of these observa- 

 tions ; viz., with an anterior separating prism to confine the grating's view to 

 one colour region, or nearly so, of the spectrum at a time. And that again 

 reduced to a minimum the chance of occasionally mistaking intruding lines 

 from an overlapping order of spectrum, for those of the order intentionally 

 in the field of view, — a source of error to which Gratings are peculiarly 

 liable. 



A ruder but more powerful source of occasionally possible error, existed in 

 an imperfect action of the inspecting telescope, combined with the peculiarities 

 of vision through a narrow vertical slit. For though when the focal position 

 of the eye-piece was too far out, that fact was easily shown, and as it should be 

 by a haziness in the image of a line, yet when it was too far in, every line in the 

 field of view split immediately and sharply into two : and these separated 

 further and further from each other, as the error of the focus increased. 

 Wherefore the appearance of close and similar double lines had to be jealously 

 watched whenever the observer's telescope, by passing from green to violet of 

 the spectrum, was innately growing in focal length. On the other hand, when 

 the definition of the atmosphere was bad, the members of a really double line 

 would throw out fringes of haze towards each other, and conceal thereby their 

 real duplicity, if very close. While all the time alteration of focus had to be 

 very sparingly employed, as it was only too apt to spoil the nicety of bisection. 



Something also of the highest accuracy may have been lost, in exchange for 

 the greater speed at which the observations both were, and imperatively 

 required to be, secured whenever the sunshine was continuous. In his 

 authoritative little work, Studies in Spectrum Analysis, Mr Norman Lockykk 

 has rightfully stigmatised the slowness of the ordinary hand and eye micro- 

 meter observation, as only enabling a careful observer " to lay down ten lines 

 an hour." But with the peculiar method, — utilising both hands and the eye, — 

 which I arranged for this spectroscope, I was enabled on Thursday, June 26, 

 1884, to lay down permanently 1865 lines in three hours. 



Yet where so many opposing difficulties are concerned, I cannot perhaps do 

 better than conform at once to the General Secretary's suggestion to give 

 further details as to the chief apparatus, and the daily circumstances under 

 which the three several spectra were measured. 



