DAMPINC-OKF IX FOREST NURSERIES. 33 



had the high value of 63 9.7 per cent. The graph indicates also a 

 decided, though less extreme, degree of variability for Peltier's 

 strains on carnations ; the survivals for the 18 strains which he used 

 in both of his experiments on cuttings have a variability coefficient 

 of 29 3. 5 per cent and the 23 strains in experiment 2 on seedlings 

 55 6.9 per cent. 



PYTH/UM 



oo 



CORT/C/UM J/rfGUM 

 '7Z7?$ t^TTP/?/ 



D D 



a en a na D a 



cPan 



OMCffftNrtTION CUTTINGS 



DDDDDan aa D an a 



PELT/E/TS STft#//VS ON CffRNflTION SEEDLINGS 



D D 



gD D D D D D 



DDDDDDD DDDD 



O SO 44 6O 60 /OO 



SK/9/V7& &/ffP/WN0 Pff /0//V CONTROLS 



FIG. 13. Diagram showing the results of inoculations with 17 strains of Corticium vagum 

 and 35 strains of Pythium dcbaryanum, arranged in decreasing order of virulence from 

 left to right, as indicated by the survivals in pots of pine seedlings artificially inoculated 

 with them. The Pythium results represent the mean survivals in 5 pots inoculated with 

 each strain in each of experiments Nos. 66, 67, and 68. Each point located is therefore 

 based on the results in 15 pots, 10 of Pinus banksiana and 5 of P. resinosa. The 

 Corticium results on pine represent 5 or 6 pots each, in two experiments (Nos. 71 and 72) 

 on P. resinosa. The outline circles represent P. debaryanum strains from East Tawas, 

 Mich. ; the solid circles represent strains from other localities. The second row of 

 squares shows the sum of the results in Peltier's experiments Nos. 1 and la (99, his 

 table 3). The lowest row of squares shows his results in experiment No. 2 (his 

 table 4). 



The strains represented in figure 13, as used on pine, include 1 

 from bean, 2 from potato, 1 from sugar beet, 1 from Elaeagnus, 2 

 from Picea engelmanni, and 10 from Pinus resinosa, P. ponderosa, 

 and P. banksiana. Two were from Washington, D. C., 2 from New 

 York, 1 from Ohio, 4 from Michigan, 4 from Minnesota, 1 from Ne- 

 braska, 2 from Kansas, and 1 from California. The sources of these 

 strains are widely distributed both as to host and locality ; they are 

 19651 Bull. 93421 - 3 



