absorption spectra of highly concentrated solutions, etc. 209 



plate was not affected to any sensible extent. It is interesting to 

 note that by using a Rowland grating and a Nernst lamp as 

 a source of light, the dispersion of the light in the third order of 

 the spectrum showed that the narrow well-defined band at A427 

 has another band adjoining it in the ultra-violet side of the 

 spectrum of not quite the same width nor the same sharpness. 

 It may be possible by using such very high dispersions to split 

 up other bands. 



With regard to the absorption bands of the solutions of 

 Erbium, the following were noticed : — 



1. A group of bands at about A 549 and A536. 



2. A very weak band at about A, 527. 



3. A very strong band at about A 5 23. 



4. A weak one at about A. 520. 



5. A broad band at about A491. 



6. A narrow strong band at about A488. 



7. A weak narrow one at about A, 48 6. 



8. Two very weak and broad ones at about A 47 2 and A, 467. 



9. A broad band at about A 454. 



10. A narrow sharp band at about A 449. 



11. A weak band at about A 441. 



12. A narrow one at about A 422. 



13. A weak band at about A418. 



14. A broad band at about A415. 



15. Two almost equally strong bands at about A 404 and 



A407. 

 A little beyond this there is no effect upon the photographic 

 plate because the light beyond is absorbed. 



Discussion of results. 



The reproductions of the photographs in Plates I and II will 

 show the differences and agreements perhaps better than any 

 description, but it may be advisable to draw attention to several 

 points in them. 



Comparing the strong and dilute solutions of Didymium 

 nitrate it is noticed that the broad band in the yellow from 

 about A590 to A570 is much broader in the stronger solution. 

 The same thing is to be observed in the band A 528 — A 520. The 

 group of two bands also in the green at about A 520 is much 

 broader and the two are really fused into one large band. The 

 bands in the blue at A483 and A469 are also slightly more 

 diffuse but not so markedly as the middle of the triplet A 47 6, 

 which is very faint and broad in the strong solution. The broad 

 faint band at A 402 seems to be much broader, as is also the case 

 with the broad well-defined band at A 444. The narrow faint 

 band at A433 almost disappears in the strong solution, whilst the 

 well-defined narrow band at A 42 7 seems to be more diffuse. The 



