24 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I40 



result joined Glock in October 1939 in an attack on problems opened 

 up by the development of methods of precise dating. Thus the work 

 has dealt with dated annual increments and in many cases with 

 growth layers dated to a part of a growing season. 



The development of the two methods of absolute dating, by the use 

 of natural and artificial frosts as time markers in the detection of an- 

 nual increments, led directly to the search for, and development of, 

 other possible methods or criteria. 



It was realized at once that a third method consisted of carefully 

 observing and measuring tip growth over a period of a year or more 

 and then of cutting sections from the measured portion of the tip 

 growth. 



As a fourth method, a discrepancy between the number of tip 

 flushes and the number of diameter flushes suggests multiplicity. For 

 instance, if two diameter flushes are incorporated in one tip flush, the 

 evidence is presumptive for more than one diemieter flush for the 

 particular growing season. A discrepancy of this kind, we have found, 

 is an entirely different type of evidence compared with the existence 

 of one diameter flush to two tip flushes, a duality far from rare in the 

 forest-border area in one season. 



The fifth method to be used in the detection of multiplicity — 

 comparison of structural features — is rather weak and has been used 

 only as corroborative evidence; growth layers of comparable width, 

 similar sequences of growth layers, and the matching of special fea- 

 tures such as a widespread injury are suggestive, but by no means 

 conclusive, time markers. 



These five methods, or criteria, are the bases upon which the entire 

 study has been elaborated. They will be taken up in further detail. 



METHODS OF ABSOLUTE DATING 



NATURAL FROST EFFECTS 



By the term "frost" we mean both frost injury and recovery ; by 

 the term "natural frost" we mean especially late (spring) frost, unless 

 otherwise stated. The typical natural frosts, which stand out prom- 

 inently in the Lubbock area and have been used almost exclusively in 

 absolute dating, are those of 1938, 1936, and 1934, in that order of 

 usefulness. For most of the work, in which natural frost served to 

 give absolute dating, that of 1938 proved to be the most typical, the 

 most uniform in occurrence, and the most readily identifiable (pis. 8; 

 19, fig. 2; 20; 24, fig. i; 29; 33). 



The original correlation of each frost with its correct year was 



