ESTABLTSHMEXT BEHAVIOR OP THE PaI.O VeRDE. 293 



metal calipers such as are used in lathe work. The fact that the 

 Palo Verde branches close to the ground made it necessary to 

 measure the trunk diameter at the surface of the ground, and in 

 some of the largest trees even to dig below the surface. When- 

 ever the trunk was elliptical or irregular in section I averaged its 

 greatest and least diameters. I then converted the set of dia- 

 meter measurements into ages, by use of the curve of growth 

 rate, and secured the following groupings of the number of in- 

 dividuals that became established during the periods indicated. 



TERIOD. NUMBER OF TREES. 



1490-1520 -- - 1 



1520-1550 - 4 



1550-1580 4 



1580-1615. - 7 



1615- 1640... ..1 1 



1 640- 1 670 1 2 



1670-1710.... 11 



1710-1760 ......18 



1760-1810 13 



1810-1 840. 1 7 



1840-1860.... 21 



1860-1885 17 



1885-1910 . 10 



It is manifest from the anti .uity of this population that I 

 have no warrant for considerin;; the above figures as any indica- 

 tion of the rate of esta1)lishment. Scores of individuals must 

 have germinated, grown to considerable size and perished on 

 this area since 1490. The ligures merely indicate the age of the 

 suiviving individuals. Howcxer, the number of survivors from 

 the 1840-1860 period to l*nO has steadily fallen, so that this 

 portion of the curve, at least, is fiee from the error of having 

 been seriously reduced by d"cath of adult trees. How much I 

 have to reckon with human interfeiencc 1 am unable to say, but 

 such interference wnild at least be directed chiefly to the taking 

 out of the laigest trees for firewood, anil might account in part 

 for the low number of smvivors of the greater ages, but would 

 scarcely affect the individuals of the last three periods. I am 



