Morphologie. 587 



mass, and, acting as a haustorium absorbs the same. The embryo 

 produces from its suspensor a number of haustoria, which, 

 penetrating between the endosperm cells, absorb food from it. 

 In Hoiistonla the integument is absent. There is there- 

 fore no micropyle. Here the embryosac develops in situ. 

 The antipodals show no special anatomical features, and the 

 embryo has no haustoria. In Diodia the supernumerary macro- 

 spores form a food conductive Strand continuous with the 

 antipodals, the number of which in D. Virginiana may be as 

 great as ten. 



(2) The tetrad divisions are found to be in accord with the 

 most recent observations. Pollen and embryo-sac divisions 

 were studied. The first division is heterotypical, the reduced 

 number of chromosomes (in Crucianella 10; in Aspenila, 12) 

 appearing in the prophases. The second division is homoty- 

 pical, the third and succeeding divisions typical. 



(3) In Diodia and Richardsonia the epidermis of the 

 strophiole forms a poUentube conductive tissue. In this, in 

 D. teres and Richardsonia, the pollentube takes an inter- 

 cellular path. In D. Virginiana however its path is superficial, 

 but is otherwise the same in all three forms. It is argued that 

 the same Stimulus determines the direction of the path in both 

 conditions, and that the Stimulus is a chemical one. The author 

 argues also for a differential distribution of the stimulant. The 

 occurence of a conductive in which the long axes of the cells 

 lie at right angles to the path of the pollentubes throws doubt 

 on Miyoshi's conclusions. Lloyd (New-York). 



Glos, D. H., La theorie du petiole dans la fleur. (Mem. 



de I'Acad. des Sc, Insc. et B.-L. de Toulouse. Ser. X. T. 1. 



1901.) 

 Le filet des etamines est d'habitude compare au petiole 

 de la feuille. L'auteur pense que c'est lä une erreur. En 

 effet: nombre de familles ont en meme temps des feuilles 

 sessiles et des filets allonges, ou inversement; dans nombre 

 d'especes Ics filets ressemblent aux petales; chez quelques 

 familles il y a passage graduel des petales aux etamines par 

 retr^cissement des premiers; les petales offrent souvent des 

 particularites d'organisation que reproduisent les filets; de nom- 

 breuses anomalies confirment l'assimilation du filet au petale; le 

 filet a la meme Organisation anatomique que le petale. En outre 

 chez nombre de polypetales diplostemones les etamines supple- 

 mentaires proviennent du dedoublement des petales; chez les 

 polypetales polystemones les nombreux filets staminaux semblent 

 correspondre aux nombreuses nervures des petales ou, au 

 moins, leurs nervures naissent du meme faisceau que Celles des 

 petales; chez les monopetales les Mens d'union entre l'androcee 

 et la corolle sont encore plus manifestes sans que rien n'y 

 revele I'existence des petioles. Rien non plus chez les Apetales 

 ne vient affirmer l'assimilation du filet au petiole, au contraire. 



