Seaton : Embryo-sac of Nymphaea advena 285 



show the floral parts, no further material was gathered in the 

 autumn, but collections were made for two consecutive seasons in 

 May, from a small lake thirty miles east of Cleveland. From 

 every collection made, buds and open flowers of varying size and 

 general appearance were selected. Although floating buds of any 

 given collection showed a wide range of gross development, very 

 little variation in the condition of the embryo-sac was found in 

 such material ; that is, size and general external appearance are no 

 definite guide to internal development in this species. The season, 

 rather than the gross appearance, is the best criterion of the con- 

 ditions of development. 



Material collected in July and August was carried to the lab- 

 oratory for killing and fixing ; that gathered in May was put in 

 the fixing fluid at the place of collection. At first, the outer floral 

 parts were removed and the pistil cut vertically into from eight to 

 twelve radial pieces, but later all except the youngest ovules were 

 removed from the pistil and fixed separately. In removing the 

 ovules, the large quantity of gelatinous substance made the work 

 imcult. Flemming's chromo-aceto-osmic mixture and chromo- 

 acetic acid were used. Because of the gelatinous substance sur- 

 rounding the ovules, Wager's alcoholic fixer was tried, but the 

 aqueous fixers gave much better results. The separate ovules, 

 w nen of small size, were slightly stained in Mo, with picro-carmine 

 after fixing, in order that they might be more easily seen during 

 subsequent treatment in grades of alcohol, cedar oil, and in par- 

 affin. Sections of varying thickness were cut and stained with 

 Afferent stains, Flemming's triple stain giving the best results. 



large number of slides was prepared and all the points figured 

 Were ob served in many preparations. 



This investigation was carried on in the botanical laboratory of 

 ornell University, under the direction of Professor Atkinson and 

 Wl th the assistance, at first, of Dr. Margaret C. Ferguson and later 

 of Dr.E.j. Durand. 



In May, before the integuments begin to develop, a single 



ypodermal archesporial cell can be distinguished {figure /). 



1S "ypodermal cell divides by a transverse wall into an upper 



ce • the primary parietal cell, and a lower one, the megaspore 



Mother-cell {figures 2, j). The primary parietal cell divides irreg- 



