88 



moves freely in the ca\-it\' thus formed in reaching the nu- 

 cellus. 



There are, in addition to the above, some further details of in- 

 terest in regard to the pollen tube. Its path, as described by a 

 number of observers for other plants, lies between the cells of the 

 conductive tissue in all cases. In Cucurbita, upon reaching the 

 base of the neck of the nucellus, it enlarges to form a " bulla " of 

 considerable size. Furthermore, this bulla in some cases pro- 

 duces lateral cul-de-sacs, which extend into the surrounding tis- 

 sues, after the manner of the lateral suspcnsor tubes in certain 

 orchids * and in the Galieae among the Rubiaceae.f It further 

 appears that the production of these lateral outgrowths of the 

 " bulla " is correlated with the supply of starch in these sur- 

 rounding tissues, since, when starch is absent, none are formed, 

 and when abundantly present, the tubes are most extensivel)^ 

 developed. Their use, therefore, according to Lorigo, is to 

 reach out after food materials, which he believes are then passed 

 on to the embryo, as a nutritive supply for its growth. 



The conclusion that there is a correlation between the growth 

 of the bulla and the presence of starch, the extent of the former 

 being in direct proportion to the presence of the latter, appears 

 to be supported by Longo's observation that, when pollen grains 

 arc grown upon gelatine, bullae are formed, while, when they 

 are allowed only moist air, they fail. These results are anal- 

 ogous to those of van Tieghem obtained for Riciinis and Synipliy- 

 fiiin. Van Tieghem, however, used gum arabic instead of gela- 

 tine. In lilodca Canadensis, however, according to R. B.Wylie, % 

 the pollen tubes which enter the ovarian cavity sometimes fail 

 to reach the micropyle, and these produce " cystoids " which are 

 usually found to be iiuitc free in the locule. W)'lic further notes 

 that these "cystoids," which, it would seem, so far as we can at 

 present sec, arc not different from the structures observed by 

 Longo and van Tieghem, are never produced in the tissues, 



^Trcul), M. Noles sur rembryogdnie de quelques Orchid<ies. Natuik. Verb. K. 

 Akad. 19 : — . 1879. 

 t Lloyd, V. 1",. op. cil. 

 J The Morphology of iE/<7f/r(; Canndemis. Hot. Gaz. 31 : 1-22. I.t 1904. 



