328 Kirkwood : Pollen-tube in Cucurbitaceae 



some other route, i. e. } traversing the tissues of the funicle and 

 integuments. Such conditions have been observed in Citcurbita 

 and AlchemillaP 



In its course from the stigma to the ovule, the pollen-tube may 

 follow a canal through the style where such is available, and inter- 

 cellular spaces as well, or it may bore through a considerable 

 amount of cellular tissue, which it may or may not affect injuri- 

 ously. Its course is usually a direct one and the direction of its 

 growth is often influenced mechanically by the structure of the 

 ovary as well as by the distribution of nutritive materials in what has 

 been called conducting-tissue. The influence of the latter as a 

 directing agent has received some attention at the hands of 

 Miyoshi 20 , 21 and others whose work we will discuss more fully 

 later, but the results from various sources go to show that in con- 

 sidering the growth of the pollen-tube we are dealing with prob- 

 lems of chemotropic irritability and the action of stimulating 

 substances whose distribution is indicated largely by anatomical 

 features. 



While the pollen-tube in angiosperms is ordinarily devoid of 

 branches, a few cases have been observed in which the tube has 

 divided into a number of branches after havine reached the ovule. 



Hofmeister 13 , describes the branching of the tubes in Pathos and 



Hippeasti 



Juglans 2 *, Hie 07 



others. It occurs also in Citcurbita as described by Longo 17 , and 

 observed by the author. That the branching is associated with 

 the function of absorption has been the accepted view, and the 

 observation of Longo that the branching is in a definite relation to 

 the presence and distribution of starch in adjacent tissues supports 

 this opinion. 



Numerous observations have been made on different plants as 

 to the time consumed by the pollen-tube in traversing the distance 

 from the stigma to the embryo-sac. This has been found to vary 

 within wide limits and to be dependent upon no fixed principle ex- 

 cept that the immediate approximation of the germ cells is corre- 

 lated with their degree of maturity. In certain herbaceous plants 

 the time required varies from eighteen hours {Limnocharis n ) to 

 several days {Crocus, Arum™) and even much longer, as in certain 



