( 605 ) 
astronomical almanac, using an adopted longitude; the parallax is 
then computed for that point of the moon’s limb, where the star 
has disappeared and which therefore has the same right ascension 
and declination as the star. We then have to add or to subtract two 
terms to or from the right ascension of the star, to get that of the 
moon’s centre, and finally we find from the almanac the Greenwich 
time corresponding with that right ascension. 
The longitude of the place of observation, then found, is the right 
one, if it agrees with the adopted longitude. If it does not agree, 
we have only to repeat a small part of the calculation with a 
modified longitude of the place, to derive the true longitude from 
the two differences. 
This method corresponds with the method, which was customary in 
the 18 century (which we find inter alia explained in the well known 
treatise of BoHNENBERGER : Anleitung zur geografischen Ortshestimmung) 
with this distinction that then the whole computation was carried 
out in longitude and latitude, whereas we use the right ascension 
and declination. Further, that for BoHNENBERGER c.s. there is no ques- 
tion of any second hypothesis. 
I will readily grant that Bessrr’s method of computing ecliptic 
phenomena and thus also for the prediction of occultations and for 
the calculation of the longitude from an observed occultation, is 
justly considered to be the classic method. It is also the only 
one explained in most of the textbooks. But it seemed to me that 
the method indicated by myself is more expeditive and only in a 
few cases inferior to that of BesseL in point of accuracy. The 
drawback of this last method consists in the troublesome preparatory 
calculations, which it requires. Any one may convince himself of 
the truth of this statement by consulting the wellknown textbook of 
Cuauvenet: A manual of spherical and practical Astronomy, Phila- 
delphia 1874, vol I, p. 550°). 
The horizontal equatorial parallax of the moon could be derived 
from the Nautical Almanac, without any correction. As for the appa- 
rent semidiameter of the moon, I myself made a determination of this 
quantity, based on an elaborate investigation in 1859, (vid. Verslagen 
en Mededeelingen der Natuurkundige Afdeeling, Vol. V1, p. 25 seqq.) 
1) I have calculated a single example by this method ; the result differed only by 
0s,1 from that obtained by the other method; in the first however 57 logarithms 
had to be taken out, against 37 in the latter. Thinking the matter over, however, 
I believe that the method of Besset will probably admit of a modification by which 
this difference will be materially diminished. I hope shortly to investigate this more 
thoroughly. 
