XXIV.] SPONTANEOUS MIXING OF TOMATOES. 399 



they show that mixing occurs spontaneously in the 

 field. It would be an interesting study in probabili- 

 ties to calculate how many plants untrue to type 

 might have appeared if all the seeds from the plants 

 had been used, instead of from nine to fifteen. In 

 all our crossing studies, it is interesting to note 

 that red varieties have never produced purple or 

 yellow fruits, while both purple and yellow fruits 

 have produced :»'ed ones. Spontaneous crossing is no 

 doubt a common means of the running out of varie- 

 ties of tomatoes. — Bulletin 33, Cornell Experiment 

 Station, 168 { October, 189 1) . 



