Connett.—On New Zealand Surveys. XXXIX 
render correct work. Lately I have adopted the principle of doing away 
with the lower tangent screw altogether, and bolting the clamping collar 
direct to the upper levelling plate. This I find to answer admirably ; the 
perfect rigidity of the lower plate thus obtained being invaluable, especially 
in trigonometrical work. 
This alteration in the construction of the instrument renders a system 
of notation slightly different from that in common use, necessary, but one 
not in the least inconvenient. : 
I shall be very happy to exhibit and explain this improvement in the 
theodolite to any practical surveyor, and also to explain the system of 
notation. 
An instrument constructed on this principle will last about four times 
as long in good observing condition as one with a lower tangent screw ; 
and instruments now useless may easily be converted and rendered again 
workable. 
In any case, I would attach considerable importance to a thorough 
periodical inspection of every instrument allowed to be used on the survey 
of the Colony. Instruments used by the Otago staff are, I am informed, 
periodically inspected, and every contract surveyor is, by the survey speci- 
fications, required to produce and exhibit the instruments to be used in 
carrying out his contract. 
Plotting Detail and other work. 
The methods of construction of maps, where the plotting of traverse 
work is concerned, pursued in connection with the English survey, I ven- 
ture to say are very rude and inaccurate; and even in laying down the 
points of a triangulation, the method used in England, and described by 
General Frome (pages 181 and 182, “ scoring triangulations” and ‘ plotting 
triangulations,”’) are, in my opinion, very far behind those in use in this 
part of the Colony. Where the scale is not a very large one, and the 
triangles are at all numerous, the methods recommended will give rise to a 
constantly increasing and intricate series of errors. ‘ 
By the system in use here of reducing all stations of the triangulation 
to the meridian and perpendicular of the initial station, the position of each 
Station is laid down independently of every other, and thus the perpetuation 
of error rendered impossible—the only really difficult problem of the plot 
being practically to raise the perpendicular accurately to the meridian. 
The results of the ordinary method of doing so by beam compasses I have 
found it useful to check and test by measuring with standard brass straight- 
edge for the proportional lengths of sides and hypothenuse of the two right 
angled triangles, which may be formed by the perpendicular on each side 
of the meridian. 
