348 smith : APPLICATION OF Till. THEORY OK 



The observations of Lock on Dendrocalamus are, as has been 

 shown generally in the preceding part of the paper, explicable 

 on our theory, but the following point also deserves notice. 

 As lias been previously stated, under conditions of high humi- 

 dity the growing culms of Dendrocalamus exude drops of 

 water from the tips and axils of the sheathing leaves. If 

 several culms of different lengths have been exposed to an 

 atmosphere of low humidity and the humidity then increases, 

 the shorter culms always begin the exudation previous to the 

 taller ones. Thus at Anuradhapura on June 19th, at 10.30 

 p.m., culm 1 about 3 feet high and culm 4 about 13 feet high 

 were both exuding drops of moisture. Culms 6 and 7 which 

 were about 22 feet high were dry at the top, but had drops 

 of water on the lower sheaths. It was not until 12 midnight 

 that these tall ones were exuding moisture right to the top. 



\ T ow since the most rapid growth permitted by the tempera- 

 ture does not take place until water is in copious supply, of 

 which the exudation of water is a sign, it follows that under 

 the usual conditions the shorter culms have a greater number 

 of hours during which their full growth is possible. It also 

 seems to follow that the higher the culms go the more depen- 

 dent their growth is upon rainfall which brings high humidity, 

 for though a short culm may reach its most rapid growth 

 while the humidity is still below saturation, a very tall one 

 will not have sufficient water supply until saturation is reached. 

 This supplies an explanation for certain facts noted by Lock 

 (12) with regard bo Dendrocalamus. Ee says "The varia- 

 tions in rate of growth become more marked in the late part 

 of the Grand Period than in the earlier, and they appear also 

 in the late portion to show a more marked agreement with 

 the nir\r of rainfall." 



W. EL de Vriea measured a flowering shoot of Agave 



amerioana and his general results are quoted by Sachs (15). 



These an not accompanied by any actual figures either of 



temperature or of the humidity of the air so that little can be 



ni»d b> discussing i hem 



