128 MEASUREMENT OF BASE LINE, 
line exactly agree; there is always a small difference, and 
though such difference may be inconsiderable, and even 
unappreciable, what is styled “‘mathematical accuracy” cannot be 
attained. The earlier measurements in such surveys were con- 
ducted under circumstances less favourable to accuracy than 
those obtaining at the presenf day. We have now the experi- 
ence gained by each successive survey, the history of which has 
been carefully recorded. It would take too long to give par- 
ticulars of the several bases that have been measured success- 
fully, although much interesting matter would be found in their 
recital. 
The chief difficulty is obtaining a unit of length, by means of 
which the measurement is made. Almost all substances, 
metallic or otherwise, expand and contract with heat and cold : 
so that it will be understood in determining the length of any 
line when measured by means of a rod, or chain, made of a 
material liable to fluctuation, it is necessary to know what its 
absolute length is at some known temperature. This tempera- 
ture is usually fixed at abcut 62° Fahrenheit. The standard of 
the Ordnance Survey in Great Britain is ten feet in length, and 
one of the same length was used in this Colony. This length is 
nominal, and is only correct when the temperature is 64°2._ The 
details of the methods used for determining the length of a 
standard would provide material for a separate paper, and I shall 
only refer to the subject by mentioning that the comparisons of 
standards often occupy months, and the greatest care is taken 
that the thermometers and micrometers used are of the best 
workmanship. At the Ordnance Survey Office, Southampton, 
is a building specially constructed for the purpose. The inner 
room, where the comparisons are made, measures twenty feet by 
‘eleven, with thick double walls, and is sunk below the level of 
the ground, and roofed with nine inches of concrete. An outer 
building entirely encloses and protects the room from external 
changes of temperature. This will givean idea of the accuracy 
