312 Royal Society :— 



occur at different places not at the same moment of absolute time, and 

 therefore the isothermal lines traced by the aid of their results alone, 

 are not true isothermal lines in the same sense as we understand 

 an isothermal line or surface within crystals, or other definite geo- 

 metrical solids which have been recently the subjects of thermological 

 inquiry. 



The distribution of sunshine at the outer limits of the atmosphere 

 and at its base is first considered, and the nearly circular shape of 

 the lines of equal sunshine is pointed out. After showing the con- 

 nexion between these lines and the simultaneous isothermals for the 

 air, land, and water, the author proceeds to discuss more particularly 

 the aerothermal lines. As the term isothermal line has become uni- 

 versal in the sense of a line joining places possessing the same mean 

 temperatures, the author proposes to designate the true lines of si- 

 multaneous equal temperature as synthermal lines. If any number 

 of places have the same temperature at a given hour corresponding 

 to the mean time of any one meridian, these places will be synthermal, 

 and a line joining them will be a synthermal line. For this purpose 

 the meridian of Greenwich has been selected, and a series of syn- 

 thermal Tables have been calculated for different places corre- 

 sponding to the Greenwich hours. For the construction of these 

 Tables, the hourly observations of temperature made at the British 

 Home and Colonial Observatories, the observations of Russia, Austria, 

 Prussia, and Central Europe, as well as those of the United States, 

 have been employed. The few series of hourly observations made 

 by Arctic and African travellers have been also applied ; and in addi- 

 tion to the Tables thus directly constructed, others have been deduced 

 by interpolation for stations whose geographical position rendered 

 it desirable to bring them into the general view of temperature-dis- 

 tribution. All results expressed in Centigrade and Reaumur degrees 

 have been reduced to the Fahrenheit scale. A fresh set of Tables 

 has been formed from those corresponding to local time, with hours 

 corresponding to the meridian at Greenwich. 



The synthermal Tables thus obtained show, as might be a priori 

 expected, still greater differences between the temperatures of places 

 in the same parallels of latitude than the Tables of mean temperature. 

 Thus Rome and Tiflis differ in latitude by only 13', and the mean 

 temperature of Rome is o°*l in excess of that of Tiflis. At 8 a.m. 

 Greenwich time, they are synthermal, both possessing the tempera- 

 ture of 59°"1, while at 7 a.m. Tiflis surpasses Rome by o, 6, and at all 

 other times besides these Rome surpasses Tiflis. At 4 a.m. this 

 excess amounts to 9°' 5. Although Pekin is situated in the isother- 

 mal line which passes close to the Isle of Wight, it is synthermal at 

 5 a.m. (Greenwich) to some place 6° warmer than Rome, and pro- 

 bably therefore on the north coast of Africa, and is synthermal 

 with a point north of the Orkneys at between 8 and 9 in the 

 evening. Similar comparisons of distant places in both hemispheres 

 lead to similar results. It appears that during certain periods of the 

 day, alternately hot and cold spaces exist in the interior of the con- 

 tinents compared to the surrounding oceans. In the southern hemi- 



