16 REPORT—1868. 
this direction of research. The chief numerical results are those obtained by 
Despretz in his well-known examination of the conductivity of water. 
In order to homologate the thermal and electrical phenomena, the term thermal 
resistance is used in preference to conducting-power or conductivity. The zero of 
thermal resistance is supposed to exist when two bodies of unequal temperature 
are in actual contact. If a substance is interposed between the hotter and colder 
bodies in such a manner that the heat can only pass between them by means of 
conduction through the interposed substance, then the difference between the 
quantity of heat which passes when the bodies are in contact and the quantity 
which passes when the third substance is interposed, is equal to the quantity of 
heat intercepted by that substance, and is a measure of its resistance. 
The instrument used for this purpose is the “ Diathermometer ;”’ its construction 
is as follows :—Two conical brass vessels having thin polished platinum bases are 
fastened in a stand in such a manner that their axes are in the same vertical 
straight line; their bases are opposed to one another and are parallel. The apices 
of both cones are made tubular. The lower cone is screwed into the stand, its 
neck is fitted with a cork and tube, which dips into water, and which carries a 
scale. The lower cone and tube form an air-thermometer. The upper cone is 
moveable vertically, its motion being commanded by a micrometer-screw, which 
is so divided as to allow of the adjustment of the cone to the 0-005 of a millimetre 
in vertical direction. The parallel bases of the two cones are adjusted horizontally 
by a spirit-level and levelling screws in the stand. The upper cone carries a cork, 
through which pass two tubes, one reaching to 0°5 millimetre of the bottom of the 
cone, the other opening just below the cork. A current of water may thus be 
made to flow through the upper cone. A large vessel of water is maintained at 
any required constant temperature by means of a thermostat. Screens intervene 
between this vessel and the diathermometer. By means of a siphon and flexible 
tubes, a current of warm water of known temperature is allowed to pass through 
the upper cone, commencing at any given moment. 
The zero or minimum resistance is found by wetting the bases of the cones 
with a little mercury and bringing them into contact (the air-film is thus excluded), 
and passing water of a known temperature for a given time through the upper 
cone. If the calibre of the tube of the lower cone is Inown, we can, by observing 
the linear depression of the column of water in it, calculate the number of heat- 
units which enter the air of the lower cone by taking into account its capacity, the 
specific heat of moist air, and the pressure to which the latter is subjected. 
If, now, the cones be separated to a known distance and a liquid be introduced 
between them, it is supported in its position by adhesion and cohesion. When 
water is passed through the upper cone under the same conditions as in the experi- 
ment when the cones were in contact, a less number of units of heat enter the air 
of the lower cone (as is shown by the smaller amount of depression in the thermo- 
meter-tube). This diminution in tre number of heat-units is called the resistance 
of the liquid under the special conditions. As the area of the base of the cones is 
known, we can calculate the resistance of a rectangular prism of known base and 
height of a liquid for a given time at a given temperature, and for a given tempe- 
rature-difference at its two extremities. 
The loss of heat suffered by the water in passing from the reservoir was esti- 
mated at all temperatures and allowed for. The absolute errors due to the absorb- 
tion and radiation of heat by the brass of the lower cone and to the accumulation 
of neated air in its upper portion, affect equally the determination of the minimum 
resistance and that of the interposed liquid ; consequently they do not affect the 
result, which is their difference. 
It was shown by measuring the time required to produce any effect on the lower 
cone, as also by interposing paper disks in the liquid between the cones, that the 
diathermancy of the liquids at the temperature-differences employed was either 
nothing or so small as to be negligible. 
Special experiments made by colouring the base of the upper cone showed that 
there was no convection. 
The resistance (measured by the number of heat-units arrested in a given time) 
of a cubic millimetre of about twenty chemically pure liquids was determined 
