748 SCIENCE. 
log. 3.816 = 0.582 
log. 4.055 = 0.608 
diff. — log. tan. = 9.974 
6 = 43° 
@ is the mean wind, and in computing it in the 
present case it was not necessary to consider the 
sign of the quadrant, as all were in the same 
quadrant. It is seen that the plotted mean and 
the computed mean agree within one degree, 
and agreements approximating to this I found 
throughout the work. Diagram 1 B, gives an 
example of another set of winds. These ex- 
amples are only to illustrate the methods. 
In the actual computations there were from 10 
to 30 observations. In my statement of the 
formula, Professor Bigelow thinks the v should 
precede the sin. 0 and cos. 0, but why he at- 
taches importance to this I do not know. The 
product of zy is the same as that of yx. It is 
distinctly stated in my paper that o is the ob- 
served direction and v the observed velocity, so 
I fail to see how there could be any misunder- 
standing ; but, if a change is needed, it would 
seem to me best to give the formula thus: 
= (v sin. 0) 
= 0 
fan = (v cos. 0) 
Having found the mean wind, my next step 
was to find the direction and velocity of the 
wind which, acting on the mean wind, would 
produce the observed wind in each case. This 
I sought to do by means of the well-known 
parallelogram of forces, as shown in Fig. 2. 
In this diagram, if 4B represents the mean 
wind and AC the wind observed at some given 
time, then AC may be assumed to be deflected 
from the mean by a wind having the velocity 
and direction of AD. By plotting the obser- 
vations in this way, I obtained the resultant 
winds for different portions of the eclipse area, 
and having plotted them, found evidence of a 
systematic circulation around the center of the 
eclipse area. This circulation followed the 
eclipse, and hence I conclude that it is caused 
by the eclipse. As before, I endeavored to 
check the plotted results and obtain greater 
accuracy by computing the resultant wind AD 
for certain of the stations. Todo this, without 
changing the position of AD, I drew a line as 
represented by the dotted line BD, and from 
[N. 8. Vou. XIII. No. 332. 
FIG. /. 
