DEEP-SEA EXPLORATION. 
333 
a square, with one side removed ami liaving tliree prongs or arms with notches at tlie 
ends that lit over tlie rim of the can. Tlie frame is thus sui)i)orted over the center of 
the vessel and on a level with its top. Two bunches of thermometers may be placed 
within the open sipiare, the sides of which are provided, in addition, with several holes 
three-eighths of an inch in diameter, through one of which is ])laced the working 
substandard, as may be most convenient for use during comparisons. A short fragment 
of small rubber tubing slipi)ed over the stem of the substandard [)reveuts it frojii 
passing entirely through the relatively large hole in the frame and permits placing 
the bulb of the substandard at the same depth in the can as those of the bnnched 
thermometers. These latter are snbmerged to within about one-half inch of the top 
of the stem. 
The two cylindrical vessels are not fastened, the one to the other. The smaller 
sets loosely but about centrally within the other. Each is titted with a dasher for 
effectually stirring the water, Avith which both vessels are tilled brimming full. These 
dashers consist of annular brass plates fitted Avith an upright rod of brass attached at 
one side and reaching just above the top edge of the vessels, resi>cctively, AAdiere a 
large thumbscrew head is attached that facilitates grasping the dasher. The dasher 
in the outer A-essel traverses the annidar space between the two; that Avithin the smaller 
vessel encircles the bunched thermometers hanging in the center. The comparator is 
placed within a shallow tray of the photographer’s tyi)e, to retain the Avater Avhich may 
accidentally slop over. 
The comparisons require the services of two jjersons — the observer and the recorder. 
The numbers of the twelve thermometers constituting a bunch are recorded in the 
record book of observations in two gronps of six numljers each, representing the six 
thermometers on one face of the bunch. An additional space in each group contains 
the number of the substandard used in the comparison, and the tAAm groups are classed 
together under a number corresponding to that stamped on the metal plate of the 
bunch; the side bearing the number is the front, the other is the back. 
The comparisons, for exanq)le at 42^, are made as folloAvs: Two bunches of ther- 
mometers and the substandard are placed in the bath. Ice in small fragments is 
added to the Avater until the temperature is less than half a degree above or below 
42°. A slight excess of ice is alloAved to remain, and if the room in which the compar- 
ison is imule is comfortably Avarm the temperature of the Avater in the outside vessel 
is regulated to be 1° or 2° lower than that of the inside vessel, in order to lessen the 
tendency of the temx)erature of the baths to rise. 
The water is stirred actively by hand, and Avhen the tem|)erature, by a feAV x)relim- 
inary readings of the substandard, is seen to change very little the stirring is suspended, 
a careful reading of the substandard is made and recorded, and immediately folloAved 
by recording the readings of the six thermometers on the face of the lirst bunch, which 
is lifted in the water a little so as to exiAose the tojAS of the columns. The bunch is 
turned back to the front, the Avater thoroughly stirred, the substandard and the six 
thermometers on the back of the bunch read and recorded. The bunch is removed 
from the Avater, Avhich is again stirred, and readings of the second bunch are made 
and recorded in j)recisely the same manner. 
When sjAecial accuracy is to be attained tlie readings at each tenqierature are 
repeated as many times as may be ilesired. 
The temperature of the Avater baths is increased by replacing a xiortion of their 
contents, withdrawn by a sixihoii, with hot water. 
