112 



MISSOURI BOTANICAL GARDEN. 



much weight. Under these circumstances I would rather 

 devote time to the collection of additional data than to the 

 mathematical analysis of those already on hand. Hence no 

 further tests for linearity of regression are applied. 



TABLE III. ALLIUM. 





















Number of Flowers 



per 



[nflorescence. 



























2^ 



00 rt 



to 

 1 



35 

 1 



1 





00 

 I 





Of 10 



M CO 





 T 



CO 

 r-H 





T 



U3 



10 00 



ia 1 iTj 



1'^ 



I 



'SI 







1 



CO 



1 



to 

 1 







01 



00 



OS 



00! o' s s 



0^ : r-l rH r^ 



to' gslN in 

 ai: Oil 



1 h -' 





 v 

 1 



1 



.g 



26-28 

 29-31 

 32-34 

 35-37 

 38 40 

 41-43 

 44-46 

 47-49 

 50-52 

 53-55 

 56-58 

 6»-61 



1 



3 



1 



1 



2 

 4 



7 



1 



5 

 2 

 4 

 2 



14 



'2 

 6 

 7 

 3 



"2 

 20 



1 

 2 



15 

 8 

 9 

 3 



1 

 1 



40 



'i 

 10 



14 

 13 

 3 



1 



42 



"3 



9 

 21 

 11 

 7 

 2 

 2 



55 



'2 



9 

 15 

 16 

 11 

 9 

 1 

 2 



65 



5 

 16 

 14 

 17 

 8 

 1 



61 



'2 



8 

 4 

 23 

 10 

 7 

 3 



57 



'3 

 5 

 10 

 9 

 10 

 1 

 2 



40 



1 



'5 

 5 

 10 

 6 

 8 

 2 



37 



■i 

 4 

 2 

 4 

 8 

 1 

 1 



21 



'3 



6 

 4 

 4 

 2 

 2 



2 

 23 



'2 

 1 



3 



4 

 1 

 1 

 1 



13 



1 

 2 

 3 



6 



i 



1 



1 

 3 

 2 



8 



1 



"i 

 1 



i 

 1 



4 



1 



2 



1 



i 



1 



;; 



i 



1 





i 



1 















i 



1 



6 

 19 

 73 



106 

 120 

 82 

 64 

 82 

 16 

 4 

 S 

 1 



1 









525 



CORRELATION. 



ALLIUM. 



Average and Probable Error 

 Standard Deviation and Prob- 

 able Error 

 Coefficient of Variation 



Length of Stalk 



39.491 ±.160 



5.446 ±.118 

 13.791 



Number of Flowers 

 per Inflorescence 



36.309 ±.340 



11.57l±.249 

 31.785 



VARIATION CONSTANTS. 



In both of these cases the regression seems to be linear, 

 although in Nothoscordum the fit of the means to the theo- 

 retical line is not very good, and in the case of Allium there 

 is some question concerning the upper part of the range, 

 where the observations are inadequate. Further tests of 

 linearity might be applied, but it seems hardly worth while 

 to do so at the present time. 



