April 1, 1920 



A Teosinte-Maize Hybrid 



31 



with either total leaves or sucker index, with the other character constant, 



is less than three times the probable error. There are also significant 



correlations with all the characters of the nodes above 



group. These correlations are in a sense physical, since 



the value representing the node silking first must always 



be greater than the nodes above. In the male branch 



group there is a significant coherence with male branch 



index and a disherence with male branches on third. In 



addition there are significant coherences with secondary 



branches, position of best spike, and days to pollen. 



DAYS TO POLLEN 



527 



^o 





y 



I 



s^ 



^^^ 

 I 



^ 



^ 



/€? 



nl 



D 



Although profoundly influenced by the environment, 

 the length of time before pollen is shed is the best 

 measure of the length of season required for develop- 

 ment. Under similar conditions there are few varieties of 

 maize that require so long a time to mature as Florida 

 teosinte, and Tom Thumb is one of the earliest varieties 

 of maize. The period for Florida teosinte under condi- 

 tions similar to those where the hybrid plants were 

 grown was 162 days, and for Tom Thumb 43 days. 



The average time for the Fj was 



98 days. The F^ plants averaged 112 



days, with a single mode at 96 days 



(fig- 31)- l^he earliest plant flowered 



U in 71 days, and the latest required 



^ ^O i — H-i 1 1 65 days from the date of planting. 



With characters of the height 

 group there are two significant coher- 

 ences, height and total leaves, and 

 two significant disherences, sucker 

 index and nodes without branches. 

 ^ The correlation with height is an especially strik- 



ing coherence, since the positive correlation is 0.47 

 while the same correlations in both teosinte and Tom 

 Thumb are negative, being 0.46 and o.ii, respec- 

 tively. Days to pollen and total leaves in teosinte 

 have a correlation of 0.14, a correlation significantly 

 Fig. 31— Days to pollen: fre- lower than the 0.79 of the hybrids. 



quency distribution of ^, ,. , ,. .,, , . , 



plants in Fj. Class value. The negative Correlation with sucker index ap- 

 10 days. pears to result from the negative correlation of 



sucker index with total leaves. 



The coherence with nodes above on third is barely significant and may 

 be physiological. There are significant coherences with three of the 

 four tassel measurements, and in Tom Thumb the three tassel measure- 

 ments recorded are all negatively correlated with days to pollen. 



Fig. 30. — Branch 

 silking first: fre- 

 quency distribu- 

 tion of plants in Fj. 

 Class value, one 

 branch. 





