=88. Sir John Herschell's Actinometer=, for ascertaining the absolute heating effect of the solar rays, in which _time_ is considered one of the elements of observation, is illustrated by fig. 68. The actinometer consists of a large cylindrical thermometer bulb, with a scale considerably lengthened, so that minute changes may be easily seen. The bulb is of transparent glass filled with a deep blue liquid, which is expanded when the rays of the sun fall direct on the bulb. To take an observation, the actinometer is placed in the shade for one minute and read off; it is then exposed for one minute to sunshine, and its indication recorded; it is finally restored to the shade, and its reading noted. The mean of the two readings in the shade, subtracted from that in the sun, gives the actual amount of expansion of the liquid produced by the sun's rays in one minute of time. For further information, see _Report of the Royal Society on Physics and Meteorology_; or _Kaemtz's Meteorology_, translated by C. V. Walker; or the _Admiralty Manual of Scientific Instructions_. CHAPTER IX. DEEP-SEA THERMOMETERS. =89. On Sixe's Principle.=--Thermometers for ascertaining the temperature of the sea at various depths are constructed to register either the maximum or minimum temperature, or both. The principle of each instrument is that of Sixe. There are very few parts of the ocean in which the temperature below is greater than at the surface, except in the Polar Seas, where it is generally found to be a few degrees warmer at considerable depths than at the surface. When the instrument is required to register only one temperature, it can be made narrower and more compact--a great advantage in sounding; and with less length of bulb and glass tube, so that the liability of error is diminished. Hence, the minimum is the most generally useful for deep-sea soundings. These thermometers must be sufficiently strong to withstand the pressure of the ocean at two or three miles of depth, where there may be a force exerted to compress them exceeding three or four hundred atmospheres (of 15 lbs. to the square inch). Many have been the contrivances for obtaining correct deep-sea indications. Thermometers and machines of various sorts have been suggested, adopted, and eventually abandoned as only approximate instruments. The principal reason for such instruments failing to give correct or reliable indications, has been that the weight or pressure on the bulbs at great depths has interfered with the correct reading of the instruments. Thermometers have been enclosed in strong water-tight cases to resist the pressure; but this contrivance has only had the tendency to retard the action, so much so as to throw a doubt on the indications obtained by the instrument so constructed.