274 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I40 



partial growth layers and so many very thin growth layers in contrast 

 with certain of their neighboring trees that cross-dating is hopeless. 



The chronology and interpretations arising from the matching or 

 cross-dating of growth-layer sequences have been accepted by some be- 

 cause they are based upon so-called "absolute dates" (Antevs, 1948, 

 p. 168) derived from trees which grew "on steep slopes underlain by 

 pervious rocks and near the lower or dry border of the forest." Thus 

 the conclusion cannot be evaded that one of two things must be true : 

 either (i) the minimum number of years represented is equal to the 

 maximum number of sharply bordered growth layers in a group of 

 specimens, or (2) the reader of rings possesses the infallible ability to 

 determine the extent of the annual increment, no matter how multiple 

 it may be or how sharply bordered the intra-annuals. 



"Absolute" dates based upon the matching (or cross-dating) of se- 

 quences from different trees, in the sense used above, scarcely ap- 

 proach the accuracy of dating in the work reported upon here. In our 

 work, something like 99 percent of extremely thin, entire growth 

 layers and lenses are intra-annual. 



The effect upon chronology of counting these thin, entire growth 

 layers and lenses as true annual increments is quite obvious. 



Insofar as a calendar year is concerned, every region has its factor 

 of multiplicity. It no doubt approaches a maximum in certain desert 

 and desert-border trees such as mesquite and oak, and in citrus at 

 Yuma, Ariz., which grows after each irrigation. It decreases some- 

 what at the extreme lower forest border and then continues to de- 

 crease toward the forest interior until it probably reaches zero. In and 

 near the lower forest border, the minimum value of the factor of mul- 

 tiplicity closely approximates the total number of extremely thin, en- 

 tire growth layers plus the number of lenses. The factor, in short, 

 depends upon habitat and species. 



MULTIPLICITY AND THE FOREST BORDER 



The character of growth-layer sequences, as well as the soil-mois- 

 ture regime, change from the desert border up into the forest interior. 

 At the extreme lower forest border, sequences are highly variable, that 

 is, they comprise a series of intermingled thick, thin, and partial 

 growth layers. The factor of multiplicity has a high value. Toward 

 the interior of the forest, variability decreases until the sequences be- 

 come uniform, that is, they include only growth layers closely ap- 

 proaching each other in thickness. At the same time the factor of mul- 

 tiplicity approaches, and may reach, zero. Soil moisture, as condi- 



