310 A STUDY OF CORRELATIONS AMONG TERRESTRIAL TEMPERATURES. 



§ 2. Irregular fluctuations tending toward a definite jteriod ; the method of time-correlation. "When the 

 fluctuation, not having a definite period, yet has a periodic tendency, this tendency can be 

 brought out by the method of time-correlation, which is also applicable to the determina- 

 tion of an unkuown period. Pp. 317-321. 



§ 3. Treatment of fluctuations without discernible "period. Criterion for deciding whether seemingly 

 irregular fluctuations of temperature in widely separated regions of the globe have any 

 common element ; method of determining this element. Pp. 321-32.'3. 



§ 4. Case when different xveights are assigned to different regions. Pp. 325-326. 



§ 5. Method of combining regions in pairs. Pp. 326-328. 



CHAPTER II. 

 Review of Data and Processes. 



§ 6. Choice and combination of material. Pp. 328-332. 



§ 7. Arrangement of the work. P. 332. 



CHAPTER III. 



Discussion of Annual Departures of Temperature. 



§ 8. Fluctuation having the period of the sun spots. Work of Koppen, — hypothesis of a harmonic 

 inequality in tempei-ature corresponding to the sun spot period — method of determining 

 the coefficients of the fluctuation — amount and formulje of the sun spot fluctuation as de- 

 termined from all readily available observations. Pp. 332-341. 



§ 9. Irregular fluctuations in the mean annual temperature. Annual departures on which the results 

 are based — evidence of correlation between annual temperatures in neighboring regions — 

 correlations of stations by pairs — small outstanding sj'stematic residuals showing a quasi- 

 periodic character, the seeming period being about six j'ears. Pp. 341-347. 

 § 10. Time correlations among annual world temperatures. A six-year periodicity strongly brought out 

 from 1871 to 1900, but not in earlier years from 1820. Pp. 347-350. 



CHAPTER IV. 



Discussion of Monthly Departures. 



§ 11. Discussion of Dove's departures. Comparison of departures of temperature in widely separated 

 regions collected by Dove — a seeming slight correlation is indicated. Pp. 351-355. 



§ 12. General discussion of monthly departures from 1S72 to 1900. A well marked correlation is found, 

 which may be attributed to the fact that some pairs of stations compared were geograph- 

 ically in each other's neighborhood — evidence of correlation in this case — omitting these, 

 some evidence of fluctuations common to the whole earth, of which the mean amplitude is 

 ± 0°.18 C. — this result is in part due to the failure to correct the temperatures for the sun 

 spot period, and partly represents the sj'stematic fluctuations of the annual departures 

 already found. Pp. 355-362. 



CHAPTER V. 



Study of Ten-day Terms. 

 § 13. Stations and material used. Pp. 363-365. 



5^ 14. Tabular exhibit of ten-day departures during the period 1871-1904- Summation of the squares 

 through annual periods. Result, absence of any coi relation whatever. Pp. 365-375. 



