480 
Selective Elimination in Staphylea 
deviation of another fruit of the same species because of the large probable error, 
but as a description of the irregularity of an individual fruit it is perfectly trust- 
worthy to as many places as we care to carry our arithmetic. I believe the four 
places to which I have tabled are arithmetically correct. For an individual fruit 
or very small sample this refinement is quite without significance, but where 
several hundreds of fruits fall in the same class, as is the case in this work, and 
especially in large series of data on other problems concerning asymmetry to be 
published later, the final decimal places may be of significance. 
TABLE XVII. 
Mean Asymmetry for Individuals. 
TABLE XVIIL 
Comparison of Asymmetry. 
Slii'ub 
A 
B 
C 
11 
•305 
•373 
•325 
12 
•488 
•456 
•4-29 
u 
•325 
•400 
■389 
15 
•478 
•470 
■424 
16 
•457 
•470 
■328 
17 
•462 
•492 
•441 
IS 
•397 
•435 
•270 
19 
•539 
■538 
•296 
20 
•644 
■494 
•374 
21 
•401 
•267 
•339 
22 
•419 
•387 
•185 
2S 
•426 
•359 
•380 
2k 
•491 
•454 
•407 
25 
•487 
•676 
•340 
26 
•500 
•470 
•359 
27 
•459 
•514 
•332 
29 
■422 
•430 
•466 
SO 
•397 
•366 
•402 
31 
•475 
•436 
•400 
32 
•372 
•360 
•287 
S3 
•435 
•491 
•417 
34 
•599 
■549 
•362 
35 
■421 
•326 
•275 
37 
•589 
■534 
•626 
38 
•285 
•361 
•339 
39 
•361 
•403 
•425 
40 
•496 
•463 
•380 
Shrub 
g ^ 
Q ]3 
C— A 
11 
•068 
— ^048 
•020 
12 
- ^032 
- ^027 
- ^059 
13 
- ^133 
•190 
•057 
14 
•075 
-•Oil 
•064 
15 
- ^008 
- ^045 
- ^053 
10 
•012 
-■142 
- ^129 
17 
•030 
- '051 
- ^021 
18 
■038 
- •les 
- ^127 
19 
- •OOl 
- ^242 
- ^243 
20 
-•150 
- ^120 
- -270 
21 
-•1.34 
- ^028 
- ^162 
-•032 
-•202 
- ^235 
23 
- ^067 
•021 
- 046 
24 
- •ose 
-•047 
-•083 
25 
•188 
- ^335 
- ^147 
- ^030 
- ■Ill 
- •Ml 
27 
•056 
- •182 
- ^126 
•008 
•036 
■043 
SO 
- ^031 
•037 
■005 
31 
- ^039 
- •ose 
- ^075 
32 
- -012 
- ^073 
-•085 
33 
•056 
- ^074 
- ^018 
34 
- •OSl 
- •187 
-•238 
35 
- -095 
- •OSl 
- ^146 
37 
•054 
•091 
-•037 
38 
•077 
- ^023 
•054 
39 
•042 
•021 
•063 
40 
- ^033 
- ^083 
- ^116 
The frequencies of the different coefficients of radial- asymmetry for each of 
the individual trees are given for the three collections in Tables XXXIV — 
XXXVI. 
The comparison may first be made on the basis of the mean asymmetry of the 
eliminated and the developing ovaries. 
Table XVII gives the mean asymmetry for the three collections from the 
28 trees. The excessive labour involved in calculating the probable errors for 
these 84 means, where the standard deviations must be calculated for classes so 
