BREEDING 



199 



have more than one flower stalk." Goff made a micro- 

 scopic study of the formation of flowers in the Clyde. ^ 

 Beginning July 5th, samples were taken weekly, both of 

 old plants wintered over from the preceding year and of 

 runners. There were no indications of flower develop- 

 ment until September 20th, when the initial stage sud- 

 denly appeared, both in the old plants and the runners. 

 By September 27th the terminal flowers and several 

 axillary flowers could be distinguished readily. "The 

 flowers appeared to start about the same time in the 

 younger rooted runner plants as in the parent plant. 

 This shows clearly that it is not age alone that deter- 

 mines the time of formation of the flowers." He explained 

 the sudden appearance of the flower buds on September 



Fig. 20. — Proliferous strawberry. 



20th in this way : " It appears that a decided decline in 



the maximum temperature occurred at almost exactly 



the time the flowers commenced to form." These obser- 



1 Kept. Wis. Exp. Sta., 1900, pp. 227-9. 



