Jory—Altitude of Celestial Object at Sea at Night-Time. 563 
necessary to consider what magnitude of error might enter into 
observations made by this method in fairly steep seas. The table 
(see p. 564) of maximum error, due to wave-elevation, will give an 
idea as to the order of the errors arising where only a partial or 
incomplete correction for wave-elevation of the signal has been 
possible. The figures given in the table are, in short, the dips due 
to half the wave-heights. In other words, supposing the signal 
was always observed when on top of a wave (or, what comes to 
the same thing, the /east altitude of the celestial object was taken), 
and no correction made for the elevation of the signal over mean 
sea-level, then the table gives the error affecting the angular 
elevation observed, and which would remain after the usual 
correction for dip in smooth water was applied to this angular 
elevation. The error is one of deficiency, and therefore the 
quantities in the table would be applied additively by way of 
correction. The table is carried so far as wave-heights of 12 feet, 
although it is probable that in such seas observation of any sort 
would be open to an equal degree of inaccuracy ; and, indeed, 
the motion of the ship and its own elevation at the moments of 
observation would render any sort of angular observation liable to 
considerable error. 
It will be seen from the table that in a swell of, say, six feet 
from crest to hollow, and in the case of an observation at a distance 
of one mile, the error, if neglected, would falsify a calculation of 
ship’s position to the extent of about one mile and a-half. Thus 
in a meridian observation of a star bearing south the distance of 
the ship to northward is over estimated by this amount. It is 
evident, then, that in ordinary weather wave-elevation need not 
introduce any serious error, and would not count against the use 
of the method under such conditions and circumstances as would 
in general render its use desirable; that is, in cireumstances where 
great accuracy was not required, but some fairly approximate idea 
of the mariner’s whereabouts was desirable or imperative. 
SCIENT. PROC. R.D.S., VOL. IX., PART Y. 2U 
