are often of service in our efforts to improve the tables by which, the movements of the moon are calculated. But this object is of very slight importance compared with those which chiefly occupy our attention during a total eclipse. The primary question in determining the astronomical value of a total eclipse relates to the duration of the phase in which the obscurity is total.
Tested by this standard, the phenomenon which we are now to consider is one of exceptional value. The phase of "totality" lasted for 4 minutes 40 seconds on the east coast of Brazil. This may seem, indeed, but a short time in which to commence and complete an elaborate series of observations and measurements; but by skilful organization of the work it is now possible for a corps of experienced observers to effect, even in this very limited time, an amount of careful work that would greatly surprise anyone who was not acquainted with the resources of modern scientific methods. Indeed, on former occasions many successful eclipse observations have been made when the period of totality has been much less than that just stated. Even in the event which we are now considering, other stations at which the duration of totality was much below the maximum have been occupied with much advantage. Thus in Chili totality lasted for 2 minutes 56 seconds. It was 9 seconds longer in Argentina. It reached the maximum for available terrestrial statistics on the east coast of Brazil; but the actual maximum duration of 4 minutes 48 seconds would be observed from a point some hundreds of miles off in the Atlantic. On the west coast of Africa, at Senegal, the duration was 4 minutes 10 seconds.
To realise the conditions under which an eclipse is
