We have already mentioned that Amon tons con structed an air thermometer which could measure temperatures as high as boiling water, which he held as a fixed point in graduation, but hitherto the importance of the points of freezing and boiling water was imperfectly estimated, and had never been united in graduating the same instrument. It was reserved for the powerful mind of Newton to put this branch of science upon the same secure footing with every other which he had studied. Had not his mind been led to more magnificent fields of investigation, he would probably have ef fected the admirable refinements in thermometric science to which he only led the way, rejecting the partial views of his predecessors, and laying the foundation, as that of every philosophical super structure should be laid, in principles as rigidly accurate in theory as they are simple and consis tent in their application.
Newton rejected spirit of wine as a thermometric fluid, as having too low a range, and selected lin seed oil as being capable of bearing unaltered in tense heat and cold. Like Boyle he computed the absolute expansion of the fluid in ten thousandths of its bulk, but he showed his judgment in not making this the basis of practical graduation, which was always to be performed by means of two fixed points, and dividing the interval into an arbi trary number of degrees. fle found that supposing the mass of oil at the temperature of melting ice to be = 10000; at the heat of the human body it was = 10256, boiling water = 10725, and of melt ing tin = 11516. As an empirical measure he di vided the space between melting snow and the common temperature of the body into 12 parts, and proceeding from the data above given, he num 725x12 bered the boiling point 256 = 34. The prin ciple of his thermometer, and a scale of various temperatures examined by it are to be found in a paper by Newton; printed anonymously in the 22d Vol. of the Philosophical Transactions in 1701, and are also partly given in the Principia.
The oil thermometer of Newton, though a happy approach to a universal standard, had striking de fects. The fluid, indeed, does not boil below 600' of Fahrenheit, nor does it freeze with a great degree of cold; its expansibility too is great, being not much less than that of alcohol; but with all these recom mendations, linseed oil has been found so viscid as to render any accuracy of observation impossible, and from the quantity which always adheres to the tube, to make the actual height very uncertain.
We now come to mention the greatest improve ment made upon the thermometer since the period of its invention,—the practical introduction of mercury as the expanding fluid. Halley had pro posed this substance, but was prevented from try ing it practically, by the consideration of its low expansible power, which indeed he underrated, making it between freezing and boiling water, instead of as has been most lately determined. This defect was certainly of consequence in the 17th century, when the art of making thermometers with bores of great tenuity was neither understood not appreciated; and for the advantageous intro duction of mercury as a fluid, a proportionate im provement in manipulation was requisite.
This practical discovery appears, as far as the testimony of contemporary writers can guide us, to be due to Olaus Romer, the ingenious discoverer of the progressive motion of light, who, according to Boerhaave,t likewise proposed the scale now known under the name of Fahrenheit's, and so ear ly as the year 1709 observed the mercury to sink by a natural cold to the zero of that scale; but it is more commonly conceived that Fahrenheit himself made that observation, and founded his scale upon it, which was contrived in the year 1720, and de scribed to the Royal Society of London in 1724.1
From this period the thermometer became of scien tific utility, and its indications being founded upon fixed principles, gave the assurance that all instru ments made with equal care would be strictly com parable. There is even now, after the lapse of a century, no prospect of materially improving the plan of the simple thermometer, or of finding any substance fitter than mercury for forming a scale of heat. In fact it unites the most important quali fications.
1st, It is found to measure with more accuracy than any fluid hitherto employed by equal spaces, equal increments and decrements of heat, which being the fundamental postulate in all thermometric fluids, is of the most primary consequence. 2d, The facility with which it is divested of air. Though this might at first sight appear a trifling recommendation, it is found in practice one of the highest value; for the difficulty with which spirit of wine is freed from particles of air is one great obstacle in forming good instruments with that fluid, the air deranging much its operations. and giving an independent degree of expansibility. Oil is still more faulty in this respect, but mercury, from its immense cohesion of parts, the high tem perature to which it may be brought, and the power of distilling it readily into a state of almost chemi cal purity, has a decided superiority over every other substance yet applied to the purpose. 3d, The range of the mercurial scale is so great, and between such limits as to be extremely convenient for all ordinary experiments; it does not boil under a temperature of of Fahrenheit according to Dalton, or according to Petit and Dulong, and it does not freeze until the extreme cold of — 39 of the same scale. 1)eluc, from whom the principal facts of the thermometric properties of mercury have been taken,* is much mistaken in asserting that mercury bears a more intense cold than alco hol. The experiments made about the middle of the last century were very erroneous in regard to extreme degrees of cold, since as mercury contracts enormously at the moment of solidification, it was supposed that a temperature of — 261 of Deluc's scale, or 489 Fahrenheit, had been measured without the freezing of mercury. On the other hand it was asserted,f that the French Academi cians had found the spirits in Reaumur's thermo meter to freeze at — 37 at Tornea in Lapland. This proves nothing but the weakness of the spirits employed, which indeed were always adulterated with water, as will be shown when we describe that thermometer. Pure alcohol has never yet been frozen by any cold, natural or artificial, though it has been reduced as low as — 91 Fah. by Mr. Walker, of Oxford, and it has even been alleged that a temperature of — has been produced.1 Foi :intense cold, therefore, such as was frequently experienced in Captain Parry's late voyages, spirit thermometers are requisite, but within all ordinary bounds, as we have already stated, mercury enjoys an ample and convenient range.