Vol. 65.] THE CALIFORNIA!* EARTHQUAKE OF ] 906. 3 



connecting points did not agree. Originally the differences were 

 adjusted as errors of observations, but in Messrs. Hayford & 

 Baldwin's report the more reasonable conclusion is adopted that 

 the discrepancies are to be attributed to displacements connected 

 with 1 the earthquake of 1868. 



As a consequence of this latest elaborate discussion of the data, 

 the numerous stations dealt with fall into three classes : first, those 

 of which the shiftings in 1868 and in 1906 are both known and can 

 be separated from each other ; secondly, those in which the latter 

 but not the former are determinable ; and thirdly, those in which 

 the total displacement connected with the earthquakes of 1868 and 

 1906 is known, but not how much was due to each separately. 



In the Report the displacement of several stations in the third 

 class is split up, that produced in 1868 being inferred from the 

 known displacements of other stations in the same region ; the 

 method is of doubtful validity, and I have thought it safer to exclude 

 the inferred displacements from consideration. 2 With this omission, 

 the data available for discussion are given in the tabular statement 

 appended to this paper (p. 15), in which the stations are re-arranged, 

 from the original list, in their order of occurrence along the fault- 

 line from south to north. The displacements attributed to 1868, 

 and the total displacements, where those of 1906 were not separately 

 determined by direct observation, are also given, but will only be 

 referred to incidentally, attention being devoted primarily to the 

 movements of 1906. 



One more explanation is necessary. In the original report many 

 of the displacements are classed as doubtful ; generally, this means 

 that the calculated displacement is so small, that it may fall 

 within the limits of errors of observation, but in a few cases the 

 doubt arises from the fact that either the number or the character of 

 the observations is not such as affords a satisfactory check. I have 

 included and used all those falling within the former category, as 

 the fact that one station moved very little, if at all, may be as 

 important as that another was displaced a considerable distance ; 

 but have excluded from the discussion all those of the second 

 category, which are distinguished by a mark of interrogation in the 

 tabular statement. 



§ 2. Turning now to the discussion of the data, we may deal 

 with the displacements in two distinct ways : either considering 

 the calculated absolute displacements, or looking merely to the 

 relative displacements as between neighbouring stations. The 

 determination of the former of these depends on the assumption 



1 I use the words connected with, as it is by no means certain that they 

 took place at the time of the earthquake, although they may reasonably be 

 regarded as the result of the same cause as that which gave rise to the earthquake. 

 In 190t>, movements near the fault-line certainly occurred at the time of the 

 earthquake; but the measured displacements of the trigonometrical stations may 

 partly have preceded and partly have followed the earthquake. 



- The inclusion of these data would not in any way invalidate, but rather 

 support, the conclusions arrived at farther on ; the support, nevertheless, would 

 be more apparent than real. 



