1870.] 



Atmospheric Lines of the Solar Spectrum, 



5 



sufficiently intense lines proves valuable for purposes of identification. I 

 may add in this place, that no lines whatever are here visible between B 

 and C ; and this fact will afford some means of estimating the relative in- 

 tensity of spectrum viewed at Mussoorie and that mapped by Kirchhoff. 



14. Direct sunlight employed. — On first commencing work, I endea- 

 voured to follow the plan adopted by Kirchhoff and employ a heliostat for 

 reflecting the sun's rays ; but, unlike the Professor, I was unable to com- 

 mand the clockwork for driving the heliostat. The variability of light 

 under these circumstances proved intolerable; add to this the necessity 

 for observing the sun as late in the evening as possible, made the introduc- 

 tion of any absorbing medium undesirable ; and, lastly, the object of main- 

 taining a constancy of circumstances between sun high and sunset, led me 

 to prefer pointing the collimator direct to the sun. I therefore screened 

 the drum and its prisms with an ample sunshade, and received the light 

 from the sun directly on to the slit of the spectroscope. 



15. Identification of constant lines. — It will be seen that in the space 

 A to C there appears but little for recognition. No doubt the following 

 groups are, generally speaking, identical, viz. : — 



My map, sun high. Kirchhoff' s map. 



Group 469-492 is the same as 470-492. 

 „ 499-508 „ „ 498-509. 

 „ 570-586 „ „ 570-587. 



But the component lines, group by group, are widely different, and several 

 sensibly intense lines in the Professor's map between 541 and 564 are ab- 

 sent here. 



16. Constant lines, continued. — In the space C to D the identification is 

 more frequent and definite. My 711 and 719 are clearly represented in 

 Kirchhoff's map, and between the latter line and 795 several other cases of 

 identity occur. Further on, towards D, the resemblance is not clear ; until 

 on arrival at the variable lines between my 955 and D recognition becomes 

 impossible. Remembering that my map has been prepared from perfectly 

 independent measures, I am naturally glad of establishing identification, 

 where possible, with the Professor's map, for the evidence such identity 

 offers of the accuracy of my work. Until my map had been finished, and 

 the scale of millimetres drawn above it, the means of making a compari- 

 son were all wanting. 



1 7. Comparison of variable and air-lines. — In discussing the variable 

 and air-lines, I much regret my inability to adopt the suggestions of the 

 Committee, and institute a comparison with Sir David Brewster's and Dr. 

 Gladstone's map given in the Philosophical Transactions for 1860. 

 The volume in question is, unfortunately, one of the few of its kind not at 

 present included in the library of the Great Trigonometrical Survey ; nor 

 am I aware that any other library in these provinces possesses a copy. The 

 only map of the kind to which I have access is that by Janssen, given in 



