8 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. Ill 



distance, north-northeast of Hohnan Pass. Unfortunately, its record 

 is short compared with that of Santa Fe, distant 40 miles to the 

 southwest from the Pass. Figure i shows the general relations. 

 Pertinent data in regard to the stations follow : 



Chacon: 7 miles north-northeast, east of divide, in the mountains. Elevation 

 8,510 feet. Length of continuous record 1 909-1 941. 



Black Lake: 19 miles north-northeast, east of the divide, in the mountains. 

 Elevation 8,348 feet. Length of continuous record 1909- 1946. 



Taos: 24 miles north-northwest, west of mountains. Elevation 6,983 feet. 

 Length of continuous record 1901-1945. 



Taos Canyon: 32 miles north, west of divide, in the mountains. Elevation 

 8,959 feet. Length of continuous record 1 909-1941. 



Las Vegas: 35.5 miles southeast by south, east of mountains. Elevation 6,400 

 feet. Length of continuous record 1887-1943. 



Santa Fe: 40 miles southwest, west of mountains. Elevation 7,013 feet. Length 

 of continuous record 1850-1944. 



Albuquerque: loi miles southwest, west of Sandia Mountains. Elevation 

 5,196 feet. Length of continuous record 1892-1946; partial record 1850-1861, 

 1863-1867, 1878-1879, and 1889-1890. 



Month-intervals chosen for the correlative tests between rainfall 

 and tree growth were: 



November-May May-July 



January-May May-August 



January-August April 



March-April May 



March- June June 



March-July July 



May-June August 



Rainfall data were then subjected to the treatment here outlined: 



1. Addition of monthly rainfall totals in order to obtain the rain- 

 fall of the intervals listed above. These sums gave raw data in inches. 



2. The raw data of each interval in inches were changed into per- 

 centages of its own mean in order to establish an identity of units 

 and an identity of base line between rainfall and tree growth. 



3. The raw percentages were smoothed by the formula 



a-f 2b-f-c 

 4 



4. The raw percentages for the stations Chacon, Black Lake, Las 

 Vegas, and Taos Canyon were merged into a group record for the 

 intervals January-August and March- July. These stations were 

 chosen because their interval averages were nearly the same. 



Method of correlation. — Because the purpose of the study was the 

 correlation of rainfall variation with growth variations and because 



