138 A. H. Graves, 



glance might have interpreted as three megaspores, the upper two 

 becoming resorbed. But careful focussing discloses another outer 

 cell at a somewhat lower plane. Murbeck (1902) figures a similar 

 case in Ruppia rostellata. 



This departure from the usual method of division of the mega- 

 spore mother-cell, in which, in general, the resulting cells are 

 formed in a straight row, is fully commented upon by Murbeck 

 (1902, p. 13), who states that it has been found also in Allionia 

 nyctaginea, Helleborus foetidus and Ceratophyllum denier sum ; and 

 to his work I refer for a fuller account of the whole matter and 

 for literature bearing upon the subject. A concise morphological 

 consideration of such a location of the potential megaspores is also 

 set forth b}- Coulter and Chamberlain (1903), who although they 

 do laot mention the case of Ruppia rostellata, note a similar arrange- 

 ment of the outer two cells as occurring in Butomus (Ward, 1880), 

 Jeffersonia (Andrews, 1895), and Potamogeton (Holferty, 1901). 



Moreover, that this position of megaspores is not an invariable 

 rule in Ruppia maritima, is shown by such a case as is illustrated 

 by PI. XII, fig. 81, where the four cells appear in a row, the two 

 outer ones already much disorganized. 



Before proceeding further, the condition shown in PL XII, fig. 82 

 should be noted, where the nucleus of the upper of the two daughter- 

 ceUs has divided, but no wall has been formed, and the whole cell, 

 along with its neighbor below, is undergoing resorption. Such a 

 happening seems natural when one reflects that the division in the 

 lower daughter-cell in general precedes that in the upper cell, 

 producing a tendency by the earlier development of the former, 

 to reduce activity in the upper daughter-cell. It wiU be seen later 

 that this omission of the wall in the upper daughter-ceU is the 

 ordinary occurrence in Potamogeton foliosus. 



In all cases the lowest of these four cells becomes the functional 

 megaspore, the upper three cells becoming resorbed, as in Ruppia 

 rostellata (Murbeck, 1902, pp. 14, 15). 



In Potamogeton natans, Holferty (19011 describes cases of four 

 megaspores with the same arrangement as in Ruppia, the innermost 

 functioning. 



In Potamogeton foliosus, described by Wiegand (1900), the con- 

 dition is quite different and yet exhibits points of similarity. The 

 first reduction division produces two daughter-cells, separated 

 by a wall. The second division then takes place in each of these 

 ceils, and the resulting nuclei occupy practically the same positions 

 as they do in Ruppia. No walls are formed after this division, 



