Hartiey— Occurrence of Cyanogen Compounds in Coal-Gas. 295 
From this point onwards to wave-length 3487°8, there are 
nineteen lines or edges of bands, only one of which is recorded 
by Eder and Valenta. They all agree fairly well with the 
bands in the spectrum of cyanogen, whether they really belong 
to this substance or not, and they agreed with no other spectrum 
which was known to me at that time :— 
Birtley, | Eder aes (Small Rgareriks. 
(| 3868 3871-4 (ON)s, D. zn 
| 8356 3855:06 (CN), K. & R. = 
| 3846 an as 
3840 3839:98 (ON)2, K. & R. se 
| 3831 383115 (CN)z, K. & R. Be 
| 3825-5 3825-4 (CN), K. &R. an 
| 3823 3823-9 (CN), K.&R. | a 
| 3818-3 3819-36 (CN)2, K. & R. we 
3815 3816:24 (ON)2, K. & R. = 
Ultra- Violet, 3790 et ee 
| 36425 | 3642-63 (ON), K&R. | { Oo A Soo 
| 3579-5 3579-22 (ON)s, K. & R. is 
| 3568-5 3568-4 (ON)z, K. & R. oz 
3563 3563-92 (ON)2, K. & R. ae 
| 3544°5 3545-07 (ON)2, K. & R. - 
| | 3528 3528-71 (ON)>. a 
| laa 22 3522-49 (CN)>. | = 
3498-5 3497-7 (ON)2. = 
3487-8 3487-61 (CN). Ee 
In conclusion, it may be admitted that a spectrum obtained 
with a prism of small dispersion is not the same as that photo- 
graphed with a grating of very great dispersion, though both may 
have their origin in the same flame. ‘The greatest differences 
between the two will appear at the less refrangible end, where 
the bands shown by the prism are resolved into groups of lines 
by the grating. It is therefore better to compare the numbers 
obtained by Eder and Valenta with small dispersion with my 
