344 Hartiey and Ramace—Banded Flame-Spectra of Metals. 
It will be noticed that several lines belonging to the spark-spectra have nearly the 
same wave-length as lines or edges of bands in the flame-spectra of the same elements; 
but the lines are not really the same, they are not lines common to both spectra. 
There is a strong continuous spectrum between the wave-lengths 5900 and 3530. 
The following are wave-lengths of the ends of the bands in Liveing and 
Dewar’s magnesium-hydrogen spectrum :—5618 to 5578, 5566 to 5526, 5518 to 
0458, 6210 to 5183, 5180 to 5112, 4849 to 4813, 4803 to 4772. 
Tue FLAME-SPECTRUM OF ZINC, 
Between wave-lengths 5900 and 3530. 
Bands composed of fine lines, degraded towards the violet. 
Wave- 
lengths. Description. 
4900 A strong continuous spectrum which becomes suddenly stronger at 4600, his 
to is distinctly seen, when suitably magnified, to be composed of fine lines very 
4600 close together. 
4811 A spark line. 4810-1 Kirchhoff. 
4722 A triplet. op ,» 4722 Huggins. 4722°37 Exner & Haschek. 
i ,», 4680 Hartley&Adeney. 4680°4 Exner& Haschek. 
4326 A broad line degraded towards the more refrangible side as far as 4320. 
4302 The less refrangi ible end of a band which fades gradually on the more refrangible 
side. 
4259 Line on a band, fairly strong. 
09) Ye oh weaker. 
50 Fe ST weakest. 
4245 A very weak line. 
4240 Less refrangible edge of a band. 
co 
~ 
———) 
Component lines of this band. A spark line at 4195°7 Exner & Haschek. 
| 
74 | 
47 
The bands composing this spectrum can be distinguished as far as the edge of 
the plate (A 8530); but the component lines of the bands are too feeble and 
nebulous to admit of accurate measurements being made. 
The lines are more widely separated as they approach towards wave-length 
3530. The spark lines of similar wave-lengths to those of lines in the flame- 
spectrum are not the same, but really different lines. 
