50 REPORTS ON THE STATE OF SCIENCE.—1914. 
Tasxe II. 
- Errors of S for Seismographs other than Milne’s. 
(Unit 0-1 m. or 6s.) 
Distance from Epicentre 
40° 80° 130° 
oO 
° 
ee 
+2-1 to 42. 5 
+1-6 to +2:0 
+1-1 to +1-5 
+0-6 to +1:0 
+0:1 to +0°5 
—04to 00 
—0-9 to —0°5 
—14to—10 | 
| —19to —15 | 
| —2-4 to —2-0 
—2°9 to —2:5 
id 
“Tho bo OO OR Re toe 
| | |. [| ere we | 
| Hronomwes | | 
se Fa 
Mean . 4 +0°5 —0°2 —0°5 
| 
_It would hereby appear that while the tables for P are fairly 
accurate, those for § are sensibly in error. The amount of error 
cannot be assigned more than very roughly by the present method, 
because the error for Pulkovo comes differently into the various earth- 
quakes. But it would appear that the times of arrival of S at 20° 
distance and at 100° distance from the epicentre are relatively erro- 
neous by something like a whole minute. The error is apparently 
not complicated in the case of § by any reflection phenomenon; the 
residuals for P are definitely grouped about two separate maxima, 
but for 5 this is not so. The first group (0°—40°) is too small to 
show a decided maximum ; but the position of the maximum is clearly 
marked in the other two by the numbers given in the table. As a 
rough expedient the following corrections have been applied :— 
Distance from Epicentre 
15° 258°! M45 bos 659 Tb? Bb" eNO abe 
Correction . —0-3 —0-2 —01 00 +01 +02 +03 +04 +05 +0-6 
the correction for 15° being applied to distances between 10° and 
20°, and so on. The corrected errors are then distributed as follows :— 
Distance from Epicentre . 
oe 2 6—— )=—40° SO —)Ss 80° — so -180°—Ss Al 
+2:8 to +3-2 te been ty | le Gee 
+2:3 to +2-7 — —_ 3 3 
41-8 to 12-2 = as aa 0 
413 to 41-7 1 1 2 4 
j +0-8 to +1-2 2 6 — 8 
40-3 to --0-7 2 4 6 12 
—0:2 to +0-2 — 11 13 24 
—0:7 to —0:3 2 2 12 16 
—1-2 to —0°8 1 4 10 15 
—1:7 to —1:3 — 2 1 3 
—2-2 to —1:8 — 1 —_— 1 
Totals 8 31 48 87 
