By the year 1700 it was just beginning to be realized that the value of meteorological observations would be vastly greater if the observations made in one place could be compared directly with those made in another. As far as the thermometer was concerned this could have been done in either of two ways. The first was that suggested, as we have seen, by Jurin: to have everyone use thermometers made by the same maker and adjusted in the same way. The second was to establish one or more scales that could be reproduced anywhere by using simple laboratory techniques; one scale, if possible, but if more than one, the readings would still be interconvertible.
Even if Hauksbee's thermometers had been as uniform as Jurin thought they were, the first scheme would have had no future, as Hooke had foreseen in 1664. The world was too large a place; too large even—up to now-for the adoption of any one thermometer scale in all countries and for all purposes. There were bound to be a large number, some better, some worse, and throughout the eighteenth century the history of the thermometer is largely that of the development of these competing scales. For all practical purposes, only three survived into the 19th century, characterized by intervals of 80°, 100°, and 180° between the freezing and the boiling points of water, and popularly associated with the names of Réaumur, Celsius,' and Fahrenheit.
There were also a large number of others, many used only by their makers, but several with a certain celebrity in restricted territories, or for one or two decades. By about the middle of the century it was not uncommon for thermometers to be made with more than a dozen scales on a wide board behind the tube. There is one with eighteen scales in the University Museum, Utrecht, dated 1754. None of the scales are centesimal. The man ufacture of such instruments went on for a long time, as witness one at the Museo Copernicano in Rome, dated 1841, and with the following eighteen scales, reading from left to right: