195 
possess the two essential conditions of rapidity and sufficient 
accuracy. It would require some little labour and time to as- 
certain from the published volumes the effectual manner in 
which the former of these conditions has been carried out. This, 
however, is easily established from the nightly noted observa- 
tions. On six occasions the nightly number of stars observed 
approached 500; on three occasions it exceeded 500, exclu- 
sively of one, when it approached 900; another 1200; and at 
last, on January 17th of the present year, no fewer than 1387 
were secured. The two last, and one of the first cited, are 
not yet published. 
«‘ With reference to the second condition, which, primd 
facie, might appear unfavourably affected by the success of the 
first, it will be seen, by reference to the Introduction already 
alluded to, that the probable error of an observation in right 
ascension of 0*:288, and in declination of 4”-27, was the result 
deduced from 1345 known stars. But this statement does not 
fairly represent the accuracy of the observations; for, be it re- 
membered, that it originates in a comparison of previously de- 
termined stars, as observed and used by us for the reduction 
of the places of those published in our volumes, and their cata- 
logued places, and with the result is mixed up the errors of 
those catalogues. Were this not the case, our probable error 
in right ascension must necessarily have appeared considerably 
less than in declination: the right ascension being obtained 
from the mean of two observations ; the declination from their 
difference. It is fair, then, to assume, that the second condi- 
tion has been fulfilled. 
« The Tables on pp.196, 197, will show, as expected, that 
meteorological records wouldfurnish no data from which a prac- 
tical astronomer could select a time for his visit to this western 
station with any certainty of securing clear nights for observa- 
tion. Were he to choose the month of May in consequence of 
its greatest freedom from rain, he would find it low in the order 
of productiveness; or November as the least variable, still the 
