110 MORPHOLOGY OP ANGIOSPERMS 



raceae, Loranthaceae, Nymphaeaceae, Polygalaceae, Lythra- 

 eeae, and Aceraceae, while they seem to Lie almost universal 

 among the Sympetalae. In most cases the advancing tip re- 

 mains narrow, but sometimes it becomes enlarged, in certain 

 cases very much so. For example, in Carina the antipodal 

 haustorium becomes a bulbous structure larger than the rest 

 of the sac, while in Saururus the narrow micropylar end and 

 the bulbous antipodal haustorium form a flask-shaped sac. 



Among the Santalaceae the vermiform haustoria sent from 

 the micropylar extremity of the sac into the cavity of the ovary 

 have been noted. Perhaps the most remarkable member of the 

 family in this regard, however, is Myzodendron, as described 

 above. In this case the haustorium is really an extreme devel- 

 opment of the antipodal extremity of the sac, but the elonga- 

 tion is so excessive that it has been included in this rather than 

 in the preceding category. Among the Fagales vermiform 

 haustoria are more or less prominent, in this case being sent 

 out laterally from near the antipodal extremity and penetrating 

 the ehalazal tissue, and being entered by the endosperm nucleus. 

 Conspicuous haustoria of this type are reported, as noted above, 

 in Fagus and Castanea, while in Carpinus the ehalazal region 

 is sometimes riddled by the haustoria from the several sacs. 

 Among the Sympetalae vermiform haustoria are common, being 

 well known among Scrophulariaceae and their allies, as well as 

 among the Campanulaceae, Lobeliaeeae, and Stylidaceae. In 

 addition to the penetration of the ehalazal tissue by haustoria 

 from the antipodal region of the sac, micropylar haustoria are 

 often sent into the tissue of the massive integument. Four 

 such micropylar haustoria, more or less prominent, and ahvavs 

 associated with active endosperm cells, seem to be common 

 among the Scrophulariaceae. Such haustoria are apt to be coe- 

 nocytie, the endosperm consisting of large and densely stain- 

 ing nuclei rather than of walled cells as in other parts of the 

 sac. The haustoria] mechanism is evident even when it con- 

 sists only of groups of active endosperm cells in contact with 

 definite regions of the sac wall. 



In this connection the remarkable case of Trapella (Peda- 

 liaeeae) may be mentioned, in which the innermost megaspore 

 of the linear tetrad becomes modified into an active haustorium 

 that penetrates the ehalazal region ( Fig. 3"2). 



