Meteorological Observations. 1 43 



from deduced, and a world of troublesome and often mistaken calculations 

 may be saved hy so applying them. 



Of the External Thermomvier. — The External Thermometer should have a 

 scale on which whole degrees are read off, and divisions large enough to ad- 

 mit of estimating tenths, or at least quarters of degrees, by the eye. It should 

 be compared with a standard, and the difference stated, at one or more tem- 

 peratures (the wider asunder the better) within the range of the climate in 

 which it is to be used. In fixing it, choose a perfectly shaded but otherwise 

 free exposure, and one where no rejkcted sunbeams from water, buildings, 

 rocks, or dry soil, can reach it : and easily accessible for reading. There fix 

 it firmly and upright. In reading it, avoid touching, breathing on, or in any 

 way warming it, by near approach of the person. The quicker the reading 

 is done the better. 



Although read off at stated times, notice should be taken of all sudden and 

 remarkable changes of temperature, as indicated by the external thermome- 

 ter, Avhenever they occur. In the neighbourhood of the Cape, and in many 

 other parts of the continent, hot winds frequently set in with great suddeiv. 

 ness, often in the night, and singular alternations of hot and cold temperature 

 occur, disturbing the regular laws of the diurnal fluctuation, and connected, 

 doubtless, with many interesting meteorological phenomena peculiar to the 

 climate of South Africa. 



Of the Maximum and Minimum, or Self-registering Thermometer^^JIhis 

 should be placed horizontally in some place out of doors, shaded from direct 

 radiation and rain, and otherwise freely exposed to air, and so fastened as to 

 allow of one end being detached from the fastening and lifted up, so as to let 

 the indexes within the boxes slide down to the ends of the fluid columns, a 

 more convenient mode, when the steel index is free enough to allow it, than 

 the use of a magnet. 



Both the thermometers should be read off as early as possible every morn- 

 ing, and the indexes re-adjusted. But as double maxima frequently, and oc- 

 casionally double minima occur, in consequence of sudden changes of tempe- 

 rature^ it is recommended occasionally to inspect both of them, with a view 

 to ascertain whether the motion of either the mercury or spirit has been re- 

 versed in an unusual manner, and such double maxima or minima, when re- 

 markable, should be recorded as " supemumerar}'," with their dates and lead- 

 ing features.* 



The Self-registering Thermometer is extremely apt to get out of order, by 

 the indexes becoming entangled in the column of fluid. In travelling they 

 should not for a moment be carried with the mercury bulb downwards ; if 

 this should happen, they are sure to arrive in a state unfit for use. To cor- 

 rect them is tedious, and always hazards fracture. With great care, how- 

 ever, it may be done, as follows : — 



lit. The Spirit Thermometer. By many jerks, force the index down to 



• The spirit thermometer is apt to undergo a gradual change of sere by the transfer 

 (by distillation) of part of its spirit to the upper end of the tub*. It ihould, therefore, 

 often be compared with the mercurial one, and the diflfereoe* of Nadangs appUed m a 

 zero. In this only case is the apphcation of a xero htfore registering permissible, and 

 indeed essential. 



