XXXVI Appendix. 
as to leave no doubt of its accuracy.” And again, “ But nothing must be 
humoured, and nothing admitted in the new record maps which does not 
plot correctly.” 
However sufficiently such a check by construction may have satisfied 
the requirements of the English Ordnance Survey, it is quite insufficient 
as a check in the detail work of this Colony, and has long been thrown 
aside in this Province as of not the slightest value, dependence being alone 
placed on the calculated traverse reductions, without which it is impossible 
to verify the accuracy of colonial surveys satisfactorily. 
The introduction of this complete system of check for detail work over 
the whole Colony, in connection with a carefully executed Minor Triangula- 
tion is, I conceive, the reform which we require. 
Improvement of Detail Surveys. 
I would venture to suggest also, that it is desirable to improve the 
character of the detail survey even in this Province, and to state my opinion 
that it is quite possible to diminish the allowed error in amount, without 
affecting the rapidity of execution, or cost of survey. 
So long as the system of chain measurements is permitted, known as 
‘plumbing the chain,” a comparatively large error may almost with cer- 
tainty be looked for. Fourteen years ago, I ceased the “ plumbing system,” 
and have ever since persevered in observing the inclination of the surface 
for each chain in ordinary undulating ground, with the best results. 
As regards accuracy, I find no difficulty in keeping the average error of 
traverse work, both chain and theodolite, under two links per mile; and I 
think, out of some four or five hundred miles of traverse now in the Otago 
Survey Office, the average error does not exceed from one to one and a half 
links per mile—the greater portion of the chain measurements having been 
taken by skilled assistants, trained on this system. 
I may say also that I find it possible to get through the work more 
rapidly by following this system than I could ever do before. 
I would therefore suggest, for the improvement of detail survey, that the 
limit of error be reduced to four links per mile, and that surveyors be 
required to observe the surface slopes, and make the necessary reductions. 
Also, that either the surveyor himself, or an educated and trained cadet, 
should make all the chain measurements. 
Particular attention should also be given to the condition and character 
of all theodolites permitted to be used in the survey of the Colony. It is 
not infrequent, in some parts of the Colony, to observe instruments in 
use which should have been for some time put on the shelf. 
Any theodolite constructed in the usual manner, after about twelve 
months’ work, begins to shew signs of “ shakiness,” and is really unfit to 
