210 Coast Survey Report for 1853. 
of particular stations, is a constantly recurring result of the sur- 
vey. They are caused by local irregularities in figure and den- 
sity of the earth and amount in several instances to about three 
seconds, while at a station of the Ordnance Survey, the station 
usual cause of these errors, though a displacement of the vertical 
to a much greater amount has in some cases been traced to this 
origin. When it is known that even now, before the mutual 
verification of sections by connecting their base lines, the tabular 
distances given in this list are generally considered as liable only 
to an average error of about one foot in six miles, it will scarcely 
seem wonderful that the station errors are found to be as distinetly 
indicated by a comparison of the azimuth and back azimuth ob- 
servations, as by those for latitudes and longitudes; so that the 
two results even verify each other quantitatively. The notes 
introducing the list give a clear insight into its mode of construc- 
tion and arrangement. 
in the very imperfect chartographic practice, now too widely 
prevalent. 
The notes present in a condensed form a classified synopsis of 
the various projections which have been used. The four classes 
into which these are distributed are based on their peculiar modes 
of mathematical genesis. The distinctive features of eighteen 
species of projections, are briefly and systematically presented. 
Bonne’s projection, being that chiefly used in Europe for top 
graphical surveys of considerable areas, is discussed in grealet 
detail. Still more space is given to the polyconic projectioDs 
which js that used in the Coast Survey office. This name 
new, and the two varieties, called rectangular and equidistant 
are both in use and require the same tables. Fortunately thesé 
methods can now be employed by any intelligent draftsman, fut 
nished with this report, in constructing any local, county, state % 
general map, within the United States. Full instrnetions 2 
given under a special head for the graphie construction © he 
rectangular and equidistant polyconie projections. The ormule 
used for computing the tables, also the constants employed @ 
their logarithms are given, though without the detailed derivatio™ 
of the formule. 4, 
The Tables are six in number. Table I. gives the relation 
between the units of length used in different conntries—Table Ht 
has for its object to facilitate the conversion into each other 
metres, yards and statute miles, and will be found highly ©” 
