66 L. D. LEET AND W. M. EWING 
The method of timing the blasts introduced further uncertainty. The 
instant of the blast, because of the failure of an electric circuit designed to 
register it on the records, was determined visually by an observer equipped 
with a chronometer. An allowance of 0.2 sec. was made for his time of reaction. 
The mean longitudinal velocity obtained from five records in the La 
Courtine experiments was 5.524 km/sec. The five values, however, varied 
so much among themselves that they were segregated into two groups, giving 
averages of 5.905 km/sec. and 5.270 km/sec. respectively. The former was 
assigned tentatively to granite, and the latter to gneiss. An examination of 
their geologic map shows that this assignment is not valid because both 
granite and gneiss were traversed in every case. 
In spite of the use of “un amortissement trés faible”, the La Courtine 
investigators report, in addition to longitudinal waves, “autres, environ deux 
fois moins rapide, d’amplitude plus grande, se manifestent sur les trois com- 
posantes; nous les désignerons par le symbole ‘L’ appliqué de maniére 
générale en sismologie aux ondes de plus grande amplitude (longues ondes). 
De plus, a La Courtine, station la plus rapprochée des explosions, ont été 
enregistrés deux fois trés faiblement sur la composante horizontale trans- 
versale, et une fois en méme temps sur la composante verticale, des mouve- 
ments trés faibles oui paraissent ainsi correspondre a des ondes a vibrations 
transversale de vitesse intermediaire entre celles dex deux categories pré- 
cedentes; nous les désignerons par ‘S’ (secondes ondes).” The use of very 
feeble damping must, of course, have made the identification of these later 
phases somewhat uncertain. In the light of experience with transverse waves 
from explosions, which has been accumulated since the completion of this 
pioneering experiment, it seems clear that the waves designated as “L” 
were actually transverse. (See, for example, refs. 3, 7). The velocity of these 
waves, as determined from the mean of eight observations, was 2800 m/sec., 
making the ratio of the velocities of longitudinal and transverse waves 1.97. 
This is practically the same as the value of 2.00 obtained in the present in- 
vestigation. 
Thus, the velocities obtained by Maurain, Eblé, and Labrouste for both 
transverse and longitudinal waves, 2.800 km/sec. and 5.524 km/sec., re- 
spectively, are about 12 percent higher than the values for Quincy granite. 
Since both the method of timing the blasts and the presence of large areas 
of gneiss in the La Courtine region could introduce systematic errors, this 
divergence is not surprising. 
Comparison of laboratory and field results 
There is a fundamental problem involved in a consideration of the utility 
of comparing small-scale laboratory determinations of rock constants with 
field results. For many years it has been customary to use constants de- 
termined in the laboratory for computing elastic-wave velocities characteris- 
tic of certain rocks. On the basis of such comparisons deductions have been 
made concerning rock types represented by velocities observed in studies of 
near earthquakes.«t*) When the magnitude of the laboratory specimens in 
210 
