38o The Ohio Naturalist. [Vol. VI, No. 1, 



to the distinct crest developed along one side. In the oldest 

 stages it was fotmd necessary to remove the hard covering of the 

 seed before cutting. Several stains were emploved among 

 which were anilin-safranin followed l>y genetian violet, Dela- 

 field's haematoxylin and Haidenhein's iron-alum-haematox\lin. 

 The latter well known stain proved to be the most successful m 

 the stages of the development of the megaspores and embr\'o- 

 sac, while Delafield's stain gave the best results with the 

 microspores. 



The flower of Sanguinaria begins its development early in the 

 summer and the ovules and stainens are considerably advanced 

 by the end of September. Material taken on Oct. ()th showed the 

 single, hypodermal, archesporial cell distinctly differentiated 

 (Fig. 2). The ovules at this time were just beginning to curve 

 1)ut no traces of the integuments were discernable (Fig.l). At 

 this time the microsporocytes were just separating (Fig. 19) and 

 a few of the tapetal cells already had two nuclei. There are 

 three intermediate layers between the epidermis and tapetum 

 (Fig. 19). Material taken on Nov. 9th showed but little change 

 in the development of the ovules, except that they had elongated 

 and the curving was much more marked. The integuments were 

 just beginning to make their appearance Ijut the archesporial 

 cell was still undivided. At this time the microsporocvtes had 

 undergone some division and some tetrads were observed. The 

 tapetal cells contained two nuclei (Fig. 20). 



By Dec. 9th the archesporial cell had divided twice giving 

 rise to the megasporocyte with two parietal cells above (Fig. 3). 

 The curvmg of the ovule was much more marked than m the 

 previous stage and the beginnings of both integmnents could be 

 readily seen in central sections. The division of the micro- 

 sporocytes must have proceeded rather rapidly for at this time 

 the micros] )ores were present with large resting nuclei and the 

 tapetum was dissolving (Fig. 21). In this stage the microspores 

 ])ass the winter. 



No material was taken from December until the first of 

 March when it seemed that the weather conditions were favor- 

 able for the continued growth of the ])lant. Sections of material 

 fixed March Ist showed that there had been but little activity 

 during the winter months. One of the parietal cells had again 

 (livicled s(j as to form a row of three cells above the large mega- 

 sporocvte (Fig. 4). In some instances transverse division had 

 taken place forming a considerable tissue (Fig. o). The integu- 

 ments had increased quite a little in size and now reached more 

 than hair way around the developing ovule. The curving of the 

 ovule was compk'te at this time. The micros]jores were in 

 ])ractica]lv the same stage in which they were found in Decem- 

 Ijer. The nucleus was undivided but the vacuole in the center 



