26 



SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I07 



If the values of normal monthly F^ and the sunspot corrections are 

 both subtracted from the mean values of Fe given in table 6, there 

 remain residuals of Fe ranging from —33 to +38 critical frequency- 

 numbers. I had hoped that these residuals might disclose valuable 

 relationships with the changes of phase and amplitude in the tempera- 



Table 7. — Ten-day and monthly means of F, 



1938. 



1941 



Jan. 



M 378 

 Feb. 



M 375 



Mar. 



M 376 

 Apr. 



M 384 

 May 



M 370 

 June 



M 342 

 July 



M 342 

 Aug. 



M 349 

 Sept. 



M 354 

 Oct. 



M 361 



Nov. 



M 370 

 Dec. 



I 



M 375 



M 346 



Jan. 



M 341 

 Feb. 



M 344 

 Mar. 



M 347 

 Apr. 



M 331 



May 



M 313 

 June 



M 311 

 July 



M 313 

 Aug. 



M336 

 Sept. 



M 344 

 Oct. 



M 353 



Nov. 



M 359 

 Dec. 



I 

 M 357 



M 334 



Jan. 



M 333 

 Feb. 



M 328 

 Mar. 



M 340 

 Apr. 



M 323 

 May 



M 303 

 June 



M 296 

 July 



M 300 

 Aug. 



M 302 

 Sept. 



M 312 

 Oct. 



M 320 



Nov. 



M 328 

 Dec. 



I 



M 329 



337 

 336 

 328 



328 

 324 

 333 



329 

 344 

 348 



314 

 325 

 330 



314 

 305 

 292 



297 

 303 

 290 



297 

 302 

 301 



300 

 307 



306 

 311 

 318 



318 

 323 

 320 



331 

 323 

 331 



337 

 320 

 331 



Jan. 



M 325 

 Feb. 



M 330 



Mar. 



M 334 

 Apr. 



M 317 



May 



M 294 

 June 



M 286 

 July 



M 287 

 Aug. 



M 293 

 Sept. 



M 299 

 Oct. 



M 304 

 Nov. 



M 314 

 Dec. 



I 

 M 322 



323 

 325 

 326 



328 

 331 

 330 



338 

 339 

 326 



328 

 311 

 313 



303 



294 

 284 



289 

 28s 

 284 



283 

 289 

 289 



293 

 300 



297 

 301 

 299 



301 

 303 

 306 



306 

 319 

 319 



321 

 422 

 324 



ture departures, associated with the 6.6456-day solar period. But 

 they do not. They do give a small positive correlation coefficient with 

 monthly solar constant numbers, but it is hardly large enough to be 

 significant. 



I have also compared the lo-day and monthly sunspot numbers 

 directly with the fluctuations of phases and amplitudes of the tem- 

 perature effects, associated with the 6.6456-day solar period. But I 

 do not find any significant correlations. 



