84 



Accordingly, at my suggestion, Mr. Chester A. Mathewson un- 

 dertook an examination of a lot of material which I had pre- 

 viousl}' collected for the purpose, and has been able to follow the 

 pollen tube from the papillae to the funicle of the ovule. A full 

 account of Mr. Mathewson's observations will appear later when 

 the work is completed. At the present it is of interest to point 

 out that the expectation entertained by me has proved correct 

 and that the course of the pollen tube is throughout intercellular. 

 Through the stylar tissue and the stylar elements of the ovarian 

 partition the tube movies precisely as described for RicJiardsonia 

 and Diodia. At the lower edge of the stylar tissue the tube 

 encounters the basal portion of the dissepiment. It then turns 

 abruptly, pursuing a path at right angles, roughly speaking, to its 

 previous course, but for only a short distance. It may turn out 

 that this is not invariably the case, though it is certainly the rule, 

 in which event the tube would penetrate into the tissue of the 

 basal element of the partition directly. Before emerging- into the 

 ovule, as it would if it kept on in the direction described, namely 

 at right angles to its stylar course, it turns again abruptly, pene- 

 trating from one to several layers deep, gradually turning so as 

 to pursue a path parallel to the axis of the placental stalk. 

 Through the parenchymatous mass of the placenta the path is 

 less direct, but in the main leads with little irregularity to one or 

 another of the ovules. On reaching one of these, the tube may 

 emerge into the sinus between the ovule and the placenta and 

 then repenetrate the ovule laterally ; or, as I believe to be the 

 more usual, the tube enters the ovule through the funicle. From' 

 this point it goes more or less obliquely and irregularly toward 

 the &^'g pole of the embryo-sac, at least in the few cases in which 

 the course has been followed. It will be of further interest to 

 see if in any instances the course is through the chalazal tissues. 

 A further question presents itself As is well known, Hous- 

 to7iia produces a goodly number of ovules in each of the two 

 locales. These are distributed upon the knob-shaped placentae, 

 which originate in a manner similar to the single ovules of the 

 Galieae and in a similar position. It seems not unlikely that 

 these ovules develop ccntrifugall}', the ones placed nearest the 



