Beer, Development of the pollen graln and anther of some Onagraceae. 287 



of primary sporogenoiis cells from tlie epidermis. The outer- 

 most parietal layer is the endothecium or fibrous layer; within 

 tJiis f'ollow (iisually) two „middle hiyers" and iinally, adjoining 

 the sporogenoiis tissue, the tapetum becomes differentiated. 



As is well known the primary sporogenous cells of Gaura 

 form a single longitudinal row. Subsequently certain of these 

 sporogenoiis cells become „sterile" and, by their division, form 

 transversa septa, here and there, across the anther. The forma- 

 tion of these septa in certain menibers of the Onagraceae has 

 already been described by Barcianii (Inaug.-Diss. Leipzig 1874 

 „Unters, über d. Blütenentwick. d. Onagraceae'"^ p. 21) and by 

 Bower (Stndies in the Morphology of spore-producing members. 

 II. Ophioglossaceae 1896, p. 1). 



Large raphide-sacs occiir in the connective of all the species 

 examined. The bündle of crystals of each sac is enveloped in 

 a mucilage which stains violet with a mixture of methylene 

 lihio and fiichsin and pink with rutheriiim red. These reactions 

 indicate a pectic body. The crystals and their mucilage sheath 

 do not tili the entii^e sac but the space which is left between 

 them and the wall of the sac is occupied by a material which 

 has often a reticulate structm-e. In Heidenhain's Iron-haema- 

 toxylin the mucilage sheath becomes black whilst the reticiilar 

 investment remains uncolom-ed (Fig. 1-1). No starch and no 

 plastids were ever seen in the sacs which enclose raphides. War- 

 ming in his description of the anther of Epilohlum angiistifolium 

 (Unters, über pollenbildende PhyUome und Kaulome. Bonn 1873. 

 p. 23) calls attention to certain large, ellipsoidal cells which lie 

 in the connective but the nature of which he left undetermined. 

 These cells ai-e the raphide-containing sacs mentioned above 

 which reach quite a remarkable development in the species of 

 EpUohium. 



In Oenotliera hiennis and 0. long'iflora tannin also occurs in 

 the anther, both in the epidermis and in a varying number of 

 cells of the •connective. On each side of the anther, along the 

 line of future dehiscence, a longitudinal band of epidermal cells 

 always remains free from tannin. The cells of these two lateral, 

 tannin-free bands soon ceases to grow and become stretched 

 and flattened by the enlargement of the anther. Beneath each 

 of these two lines of peculiar epidermal cells a longitudinal air- 

 passage is formed at a very early stage^). This passage arises, 

 in the first place, by a Separation of cells from one another at these 

 Spots but subsequently the cavity is enlarged by the cells bor- 

 dering lipon the space becoming flattened and destroyed by the 

 growth of the anther. 



In some anthers a curious develoi)ment of the cell-walls 

 bordering upon the air-passage was observed. The cell-walls in 

 question become greatly thickened and cuticularised in a manner 



1) Some time before the appearance of the callose mother-cell walls. 



