ON SEISMOLOGICAL INVESTIGATIONS. 57 
provisionally extended. It seems clear that an even larger correction is 
necessary at this distance than has been assumed. In calculating 
the mean error these observations have been omitted, and the mean 
error is then +1:1m. as before. Including them as they stand 
raises if to +1°2m. : 
TasLe VII. 
36 Errors for S for Miine Seismographs in 1911. 
| 
Distance from Epicentre in kms. 
0 — 5000 — 9000 — 10000 — 11000 — 16000 — All 
+2-8 to +3-2 
+2-3 to +2-7 
+1:8 to +2-2 
+1:3 to +1-7 
+0-:8 to +1-2 | 
40:3 to +0-7 
—0-2 to +0-2 
—0-:7 to —0:3 | 
—1-2 to —0:8 
—1-7 to —1:3 
—2-2 to —1-8 
—2-7 to —2-3 
[Pe Stiniec Ses ate 
—~- 
hoe 
Sr 
Te SN 
Re DARRPOWOND eS 
1 
1 
1 
1 
Per rwro| | He | 
LPs 1 sledge hare en] 
In addition there are three large positive errors (+9°9m., +7°8m. 
and +7Sm.) and four large negative (-52m., —5'8m., —6'7m. 
and —81 m.), which may be reflected waves. The percentage is 
slightly less than before, but, putting 1911 and 1918 together, we 
have 36+39=75 tolerably certain S readings as against 47+25=72, 
or possibly 644+39=103 P readings. The fact that S is as often 
readable as P on Milne seismograms, at any rate for large earthquakes, 
seems to be thus fairly well established. 
VI.—Comparison of Milne and Galitzin Instruments. 
To the information conveyed by the above discussion the following 
may be added. At Eskdalemuir Observatory various seismographs have 
been mounted side by side for comparison, and Mr. G. W. Walker 
made very careful and thorough comparisons of the relative advan- 
tages as indicated in his book already referred to. It seemed desirable 
at the present juncture to have a formal report on the comparison of 
the Milne instrument with at least one other; and the Galitzin seemed 
the best to select as standard of comparison. Application was there- 
fore made to the Superintendent of the Meteorological Office, and he 
kindly sent the following report, to which the names of L. F. 
Richardson and L. H. G. Dines are attached. 
Comparison between the Milne and the Galitzin types of Seismographs. 
It is convenient to treat the question under several different 
aspects, and a brief description of the two instruments may usefully 
precede the rest. 
It is unnecessary to say much about the Milne instrument. 
Extreme lightness and compactness characterise it, and no simpler 
