1918] JONES—HETEROSIS 327 
and crossed seeds in a reciprocal combination on two ears is shown 
in table I.t 
TABLE I 
DISTRIBUTION OF WEIGHTS OF SELFED AND aac: SEEDS OF MAIZE GROWN IN 
SAME INFLORESCENCE 
eae b 4 4 ead Weight of seeds in Total Per- 
ant number; seeds ondi- centigrams oO cent- 
wn in same in- — of | tion of |__- num- — Increase | age 
orescence . seeds ber 18 in- 
10/14/18/22|26/30/34/38/42 crease 
14-10-4-6-4-—7-26..| Yellow | Selfed I] 7|53|11| 1 73 |30.20.19 
14-10-4-6-4-7-26 x Light 
20A-—4-25-37..... yellow | Crossed I] 0] 3}12/72|90\10| 188 |[35.9%0.16|5.7+0.25| 18.9 
20A-—4-25-36....... White | Selfed | 1| 1| 2/63] 2 69 [21.70.16 
20A-4-25-36X14- | Light 
10-4-6-4-7-6....| yellow | Crossed 6/33) 5 44 |25.9*#0.20/4.20.26) 19.4 
The crossed and selfed seeds on one of the ears shown in the 
table differ by 5.7 cgm. in average weight, a divergence which is 
22 times the probable error. The reciprocally crossed ear produced 
seeds which differ by 4.2 cgm., or 16 times the probable error. 
One ear with 5 crossed seeds and 328 selfed seeds gave the 
largest increase obtained in all the pollinations. The selfed seeds 
altogether averaged 37.3 cgm. in weight, while the 5 crossed seeds 
averaged 58.0 cgm. This is an increase of 55 per cent. Among 
the selfed seeds, however, were all the tip seeds, which were smaller 
in this ear (as is nearly always the case in maize) than the other 
seeds. The comparison is therefore unfair to the selfed seeds. 
Taking only the 10 seeds immediately adjacent to the 5 crossed 
seeds on the basal and apical sides the increase was still the largest 
obtained, 35 per cent. The crossed seeds were visibly larger, as 
shown in fig. 3. 
The fact that the greatest increase was obtained where the 
proportion of crossed to selfed seeds was least, suggested that the 
heterotic seeds developed at the expense of the selfed seeds. An 
examination of all the data, however, showed that there was no 
Significant correlation between the amount of increase and the 
* A Jolly balance was fitted with scale and pointer so that the weights could be 
read off directly. A pan was constructed out of stiff paper in such a way that pressing 
the two ends together allowed the seeds to fall out through a slit in the bottom after 
weighing. This proved to be a great time saver. A magnifying glass helped in read- 
ing the scale (fig. 2). 
