Studies on Pollen — IV 265 



32 annual rings for the total number of years of the tree's age, and 

 these stood out with unusual clearness. In addition, however, during 

 the two years of defoliation above referred to, three other rings were 

 observable although somewhat less clearly. A reference to the Weather 

 Report records gave the explanation of these narrower annual rings. 

 When the tree was 15 years old there occurred after the leaves were 

 formed frosts on two successive nights vs^hich were so severe as to 

 cause defoliation. After this leaves were again formed but later in the 

 same year were destroyed by insects, a thing which rarely occurs with 

 Liriodendron tulipifera. These two defoliations caused the formation 

 of two rings or midsummer growths in the tree in its fifteenth year, 

 which were not so distinct as the usual annual rings. The formation 

 of two rings in one season is exceptional, but it shows, as stated by 

 Jost, that a relation exists between defoliation and annual ring forma- 

 tion. When the tree was 23 years old it was again defoliated early in 

 the season by frost which caused an extra ring to form. This was less 

 distinct than the annual rings. These same midsummer growth forma- 

 tions were visible in some maple trees that were in the same locality, but 

 they were less distinct than those mentioned for Liriodendron. 



Various and numerous theories have been advanced concerning an- 

 nual ring formation but as yet the real causes are unknown. Mechani- 

 cal explanations have failed but the teleological viewpoint deserves study. 

 While some facts are known concerning annual ring formation in tem- 

 perate zones extremely little is known concerning annual ring formation 

 in tropical trees. The first case of growth rings on record in a monocot 

 has recently been reported by Chamberlain' for Agave fcrox which 

 grows in South Africa. In view therefore of the production of the mid- 

 summer growths as recorded in this paper, controlled experiments, al- 

 though they would be costly to perform, should be carried out in order 

 to determine the physiological facts involved in this important question. 



STUDIES ON POLLEN— IV. 



F. M. Andrews, Indiana University. 



Since my third contribution to this subject I have investigated the 

 germinal behavior of more than a score of different pollens not hereto- 

 fore considered in this series of studies. The more one progresses in this 

 line of investigation, as in many others, the more one finds to be done 

 in the various interesting and important phases of the subject. It is 

 also clear that in many respects corrections of previous work needs at- 

 tention and re-investigation from different angles. Improvements in 

 the methods of research in this subject are constantly required as the 

 work has progressed. The petri dish method has practically supplanted 

 the procedures usually employed. A temperature arrangement has been 

 devised which has definitely advanced this field of investigation. Too 

 little attention has been paid heretofore to conditions controlling pollen 



1 Chamberlain, C. J. Growth Rings in a Monocotyl. Bot. Gaz., 72, 293-304, 1921. 



