ME. J. PRESTW1CH ON SUBMARINE TEMPERATURES. 
611 
the corrections necessary for his several observations. He showed that a variety of consi- 
derations have to be taken into account with Hales’s water-bucket or any similar appa- 
ratus, and that the scale of corrections must vary with the latitude and the depth. 
Thus in lat. 21° 14' N., with a surface-temperature of 79°-5 F. and at a depth of 2635 feet, 
his corrections amounted to 4° F., while in lat. 45° 53', with a surface-temperature of 58°*3 
and at a depth of 2524 feet, they amounted only to 0 O- 6 F., and, again, for the lesser 
depths of 898 and 1252 feet in the same latitude respectively to 0 O- 4 F. and to 0 o, 6 F. 
The same corrections cannot, however, be applied to the observations of Ellis, Cook ? 
Forster, Irving, Scoresby, Franklin, and Wauchope ; for in the case of the first three 
and of Franklin the apparatus was not protected by any other non-conducting sub- 
stances ; in the case of Wauchope’s and Scoresby’s later experiments the correction 
must be applied to the enclosed Six thermometer ; and in Irving’s the small size o^ 
the apparatus, although protected, necessitates a larger correction. It is, nevertheless, 
satisfactory to note, from the regular decrease in the value of the corrections from the 
equator to the pole, that in the higher latitudes, where Hales’s apparatus has been most 
used, the special corrections needed for that apparatus diminish to their minimum, and 
are so small that probably 0 o, 5 to 1° would cover all the errors of observation made by 
the foregoing explorers. The main error for correction is that due to pressure in those 
instances where a Six’s thermometer has been used in conjunction with Hales’s appa- 
ratus. 
The second plan, that of sinking an ordinary thermometer, protected and surrounded 
by some substances which are bad conductors, has been but little used, as it requires so 
much time. Independently of this, and for moderate depths, it is trustworthy and 
useful, and some of the results, as those of Saussure, may be accepted as closely accurate. 
The third plan, that of taking the temperature of mud or silt brought up from the 
bottom, has the advantage that it secures the possession of a body having the exact 
bottom-temperature ; but it has the disadvantage of small bulk, and therefore of being 
more influenced by the temperature of the water through which it has to pass. For 
moderate depths, however, the error can only be small. 
The first and last of these methods, whatever their inconveniences, had but one main 
source of error — causing a gain where the surface-temperature is higher, and a loss 
where lower, than that at depths. Only in one instance, however, was the necessary 
correction accurately estimated. But with the introduction of the self-registering ther- 
mometer two sources of error (the one occasional and uncertain in amount, arising from 
shifting of the indices; and the other fixed and definite, resulting from pressure) 
were introduced. Owing also to the want of standard instruments, the observations 
made on the several voyages have had in themselves different degrees of value, dependent 
on the care with which the instruments were made, and on the precautions with which 
they were used. As such precautions were, it is evident, usually enforced, and Admi- 
ralty instruments were generally used, a considerable uniformity of result has been 
nevertheless maintained ; and the readings on the different voyages agree sufficiently 
MDCCCLXXV. 4 N 
