5i6 ANGIOSPERMAEMONOCOTYLEDONES 



xanthum, Pennisetum, and Spartina). The anthers empty most of the pollen at once 

 by turning completely over. When the stigmas project laterally from the pendulous 

 or nodding spikelet, they bend upwards and are thus only dusted with the pollen of 

 flowers situated above. Rarely they project from the tip of the spikelet ; this occurs 

 in protogynous and monoecious species. Cleistogamy is not infrequently found, 

 according to KiefFer, e. g. in Leersia oryzoides Sw., Festuca Myuros Z., F. bromoides 

 Z., F. ciliata Pers., and others (Bui. soc. bot., Lyon, viii, 1890). 



This process is not seen in other genera. The glumes in species of Alopecurus, 

 Anthoxanthum, Mimbora (verna), Crypsis, Nardus (stricta), Phalaris, and Phleum, 

 open scarcely or not at all during anthesis. In these cases stigma and anthers project 

 outwards through a narrow slit. Phleum and Phalaris possess vestigial lodicules ; 

 in the other genera named they are quite absent. 



Ludwig (Bot. Centralbl, Cassel, viii, 1881, p. 87) says that the succulent, shining 

 lodicules of many grasses, serving as expanding bodies, sometimes seem to attract 

 flies, which are imprisoned by the rapid closing of the glumes. He repeatedly 

 observed flies imprisoned in this var., and already partly destroyed, on Molinia 

 caerulea Moench ; these had all been caught by the proboscis, which was firmly 

 gripped by the glumes below the lodicules. 



Ludwig (Bot. Centralbl., Cassel, xviii, 1884, p. 123) subsequently found his 

 supposition only partially confirmed. At Greiz he observed thousands of hover-flies 

 (sp. of Melithreptus, Melanostoma, Platycheirus) on Molinia cz.tr\AQ2i Moench, most 

 of them dead and swollen out of shape, some still living, but suffering from the 

 Entomophthora disease. On the flowers (with blue anthers) of Phleum pratense Z., 

 Avena pubescens Huds., Dactylis glomerata Z. (and Plantago lanceolata), there were 

 also flies suff"ering from this fungoid disease, but much more rarely than on Molinia. 

 A large number were adhering to the flowers (many being fastened to the anthers in 

 Molinia), but still more were trapped in the manner described above. The flies observed 

 earlier seemed to be quite free from fungus. In the cases noticed later it may have 

 been the result of the fungoid disease that the Molinia was so closely covered with 

 flies. Their frequent adherence gave the impression that they had been compelled 

 by thirst to seek the juice of the lodicules, and overtaken by death while sucking it, 

 the proboscis being then fastened in later as the outer glume closed. 



Sprengel ('Entd. Geh.,' pp. 26, 79-80) evidently also observed the lodicules, as 

 he speaks of the 'nectar-glands' of the grasses; the apparent presence of nectar 

 in flowers of grasses, which otherwise possess all the distinguishing marks of 

 anemophily, was a problem which he could not solve. 



Many grasses are protogynous, so that self-pollination is frequently prevented ; 

 automatic self-pollination does, however, often take place, and cleistogamous flowers 

 are also tolerably frequent (e. g. in Oryza, Stipa, Bromus, Hordeum, Cryptostachys, 

 and others. 91 



Kiefer states that Festuca Myuros Z., F. bromoides Z., and F. ciliata Pers. bear 

 cleistogamous flowers. 



Andersson and Hesselman ('Bidrag till Kanned. om Spetsbergens o. Beeren 

 Eil. Karlvaxtflora,' pp. 70-8) have determined more or less accurately the flowering 

 and fruiting seasons of the following species : Festuca rubra Z. var. arenaria Osb., 

 F. ovina Z. vars. violacea Nath. and vivipara Z., Glyceria angustata T. Fries, 



