118 



SCIENCE. 



IN. S. Vol. XX. No. 490. 



The second diagrani represents the air- 

 bubbles as gradually narrowing upwards. This 

 accords with the theorem of Schwendener and 

 Steinbrink, who held that the duets are ex- 



SeptTXTTb. 



SeTTvi - s eptxiTTV. 



Xylem-dxict. 



(diagraTUTnatic) 



ceedingly narrow at their tips. The same 

 result is given from observations on the red 

 beech by Hartig and Weber. Strasburger had 

 previously shown that narrow ducts contain 

 very little air, and have streaming water; 

 whilst large ducts in tall stems have much air 

 and usually little water. If we accept these 



data we find an extraordinary correlation. 

 Through the length of a lofty stem are wide 

 tubes, whose contents are a column of froth 

 of the lightest kind, having a maximum of 

 air in a very thin shell of water. In the 

 region of the leafy spray the conditions are 

 reversed, narrow ducts and a relatively heavy 

 load; and hence the need of a high vacuum, 

 which is secured by the curious structure and 

 the proximity of the leaves. 



I am obliged to Professor MacDougal for 

 referring me to Steinbrink's paper, and to 

 Mr. Earle Anderson for drawing the diagrams. 



George MACLOsjiiE. 



Princeton Univeesitt, 

 June 25, 1904. 



BOTANICAL NOTES. 



THE NUMBER AND WEIGHT OF COTTONAVOOD SEEDS. 



At my suggestion, one of my students, Mr. 

 B. E. H. d'Ailemand, made careful counts 

 and estimates as to the number and weight 

 of the seeds of the cottonwood {Populus del- 

 toides). Selecting a well-grown pistillate 

 tree about forty feet in height with a trunk 

 two feet in diameter, and a spreading top 

 fully forty-five feet from side to side, he care- 

 fidly divided it by an imaginary vertical 

 plane into two equal parts. One of these 

 halves he divided again in the same manner, 

 and continued the process until he reached a 

 branch small enough to enable him to count 

 the number of catkins which it bore. It was 

 found in this way that the tree bore about 

 32,400 catkins. Then a number of careful 

 counts were made of the seed pods in the cat- 

 kins, by which it was found that the average 

 number is about twenty-seven. The average 

 number of seeds in the pods was easily de- 

 termined by a series of counts to be thirty- 

 two. From this it appears that this partic- 

 ular tree produced the enormous number of 

 nearly twenty-eight millions of seeds. 



One hundred seeds with their cottony fibers 

 attached were then weighed upon a chemical 

 balance. The result was .065 gram. So the 

 weight of a single seed is .00065 gram, and 

 the total weight of all the seeds on the tree 

 18.3 kilograms, or almost exactly forty pounds. 



