268 Proceedings of the Royal Society of Edinburgh. [Sess. 
no necessity for any readjustment whatever. In actual work such facilities 
are of considerable importance. The two Nernst lamps are run in parallel 
off the town mains, but as both are placed in the same circuit it is hoped 
that small fluctuations of voltage may affect each equally. 
§ 3. Results. 
It is not proposed in the present paper to give a complete discussion of 
the following results; indeed, it has been found that various points of 
unexpected difficulty have been raised, the consideration of which must be 
postponed for the present. Meanwhile, however, certain remarks fall to 
be made. 
Table I. 
Photo- 
micrographs. 
Mean Wave- 
Plotted 
No. 
Nature of Scattering Bodies. 
length of 
Light. 
in 
Figure. 
1. Light Incident Normally. 
Plate. 
Fig. 
1. 
Plate glass, one side very roughly ground, 
I. 
1 
5876 
2 
ground side towards light. 
2. 
The same, but ground side away from light. 
I. 
1 
5876 
2 
3. 
Two pieces of ordinary commercial ground 
I. 
2 
6678 
3 
glass, ground sides together. 
4. 
The same. 
I. 
2 
5876 
3 
5. 
The same. 
I. 
2 
5016 
3 
6. 
Specially finely ground glass, ground side 
I. 
3 
5876 
4 
towards light. 
7. 
The same, but ground side away from light. 
I. 
3 
5876 
4 
8. 
Ground glass etched by acid, ground side 
I. 
4 
5876 
4 
towards light. 
II. 
5 
9. 
Turbid solution of resin. 
6678 
3 
10. 
The same. 
5016 
3 
2. Light Incident Obliquely. 
11. 
As in 6. 
I. 
3 
5876 
5 
12. 
As in 7. 
I. 
3 
5876 
5 
13. 
Kepetition of 11. 
I. 
3 
5876 
5 
3. Theoretical Gases. 
Glass plate with cavities all over the face 
which is next the light. 
Kefractive 
index 
of Glass. 
14. 
Semicylindrical cavities. 
1-5 
6 
15. 
Hemispherical cavities. 
1-5 
6 
