90 TUB SAPEOLEGNIACEJE OF THE UNITED STATES, 



are absorbed into the masses, which we may call, following the homology of the spo- 

 rangium, oosphere-origins (Fig. 26, h). If the rapidly shifting vacuoles present in 

 the sporangium at this stage are also formed here, the protoplasm is too dense to per- 

 mit their recognition. The oosphere origins, which, when numerous, may nearly fill 

 the oogonium at their period of greatest swelling, now contract rather rapidly, ap- 

 proaching more and more nearly to the spherical form. During this process there are 

 separated from the origins small masses of protoplasm which may move away a short 

 distance and may remain detached for some minutes (Fig. 26, c) ; but they appear to 

 be always taken up again by the same origins from which they were separated (Fig. 

 26, d). The rounding off is soon completed, and the oogonium contains a number of 

 fully formed o'dsphercs. All the oogonia of some species, and the smallest of most 

 others, produce only a single oosphere in each. The formation of these follows the same 

 course as that above described for the polysporic oogonium, with certain necessary 

 simplifications. As has been intimated, the oospheres are normally spherical, but they 

 may assume an ellipsoidal or cylindrical form when compelled to do so by the size 

 and shape of the space within which they are developed. 



The separation of protoplasmic fragments from the zoospores and oiJspheres during 

 their final contraction and rounding off, and their subsequent reabsorption by their 

 parent masses, constitute phenomena of peculiar interest. They were first observed by 

 DeBary ('81) in connection with the oospheres, where they are the more conspicuous; 

 and their formation has been regarded as analogous with that of the polar bodies of 

 the animal egg, while their reabsorption has been explained as compensating for the 

 absence of an act of fertilization. But the fact that the nuclei of the oospheres are 

 reduced to one or a very few at the time of their formation makes it certain that these 

 fragments are non-nucleate and therefore not analogous to polar bodies ; while the 

 fact of their separation from the zoospores also removes the possibility of their sexual 

 significance. They probably represent in their formation the persistence of some 

 inherited phenomenon of no present functional significance, as Hartog ('92) has sug- 

 gested. 



It will be seen that, omitting the preliminary formation of vacuoles, the changes 

 which characterize the formation of the oospheres are identical with those observed 

 in the development of the zoospores. But the time required for the former is much 

 longer, for the zoospores may escape an hour after the cutting off of the sporangium, 

 while the oospheres may require eight hours or more for their formation. 



As soon as the oospheres are differentiated, the antheridia, when present, begin to 

 produce the fertilization tubes, which soon reach and come into more or less close 

 contact with the former. The tendency of the tubes to grow towards the oiispheres 



