CORRELATIONS IN NOTHOSCORDUM AND ALLIUM. 107 



tween length of stalk and number of flowers per inflores- 

 cence is shown in Table I, while that from the southeast 

 slope numbered 426 plants whose characteristics are shown 

 in Table II. 



TABLE II. NOTHOSCORDUM. 













Length of Stalk in Centimeters. 



















10 



11 



12 



13 



14 



15 



16 



17 



18 



19 



20 



21 



22 



23 



24 



25 



26 



27 



28 



29 



Totals 



a 

 a 



2 



4) 



Ft 



3 

 4 

 5 

 6 

 7 

 8 

 9 



1 



i 



2 

 1 



2 

 1 



1 



3 

 6 

 3 

 1 

 2 



2 

 3 

 8 

 2 



4 

 10 

 8 

 3 

 2 



2 



15 



21 



4 



1 



2 



7 



22 



7 



5 



2 



11 



22 



6 



6 



1 

 9 



22 

 9 



10 

 1 

 1 



i 



16 



12 



7 



3 



h 

 11 



13 

 6 

 3 



i 



9 

 6 

 8 

 3 



27 



3 

 6 



7 

 10 



1 

 1 



28 



i 



5 



1 

 8 



1 



16 



2 

 2 

 3 

 1 



8 



i 



1 



i 



2 

 5 



3 

 3 



i 



7 



i 

 1 



i 



2 



"i 



4 



i 



19 

 74 

 159 

 82 

 72 

 14 

 6 



o 



Totals 



2 



715 



15 



27 



43 



43 



47 53 39 



1 1 



38 



1 



426 



CORRELATION IN SECOND COLLECTION, SOUTHEAST SLOPE. 



In these tables the numbers in the left marginal column 

 represent flowers per inflorescence while those along the top 

 give the length of stalk to the nearest centimeter. 



A word must be said concerning the method of collecting 

 the samples. To dig out the bulbs for so many individuals 

 would have been an almost impossible task.' By pulling 

 steadily and gently on the stalk it breaks a considerable dis- 

 tance under ground and at about the same point near the 

 bulb. It is fairly easy to detect stalks which have broken 

 sensibly above the usual point and such were discarded. 

 This method of gathering the stalks is a source of some error. 

 The means are necessarily slightly reduced below what they 

 would have been could the bulbs have been dug out. Pos- 

 sibly, too, the variabilities are slightly increased. But the 

 purpose of this first investigation was merely to get some 

 general idea of the degree of interdependence of length of 

 peduncle and number of flowers per inflorescence, and for 

 this purpose very precise measurements are superfluous. I 

 think the error introduced by the method of collecting the 

 plants is— for the purposes in hand— so slight that it may 

 be neglected. At the time the material was gathered prac- 



