STUHV OF THE GENICULA OF CORAM. l S.V.. 23 



cells also stains clearly. Its thickness, however, is far less than thai of 

 the extragenicular portion. Cheil. anceps, Amp. tuberculosa and several 

 species of Comllina have the periclinal cells running longitudinally but 

 undulating; and their zonal arrangement is much disturbed (fig. 4). 

 In these plant« there are spaceous rooms between the cells whenever the 

 latter become parted. The middle lamella does not fill up these rooms 

 and intercellular spaces in an exact sense are met with. 



The cellwall of the periclinal cells at the medulla stains in some 

 degree with ruthenium red, but those of the cortical cells in most 

 <as«s remain unstained. In Amp. tuberculosa the genicular portion 

 ] »roper stained in only a very slight degree. But in most others the 

 portion stains as deep as the cellwall of the medullary cells of the 

 articulus. The cells at the circumgenicular cortex with its epidermal 

 cells also have their cellwall stained in the deepest degree. 



The genicula of Amp. dilatata, Amp. ephedrœa, Amp. Bowerbankii, 

 Amp. echigoensis, etc., probably all of the Eurythn, are built up with 

 several zones of periclinal cells. I would call them multizonal 

 genicula to distinguish them from the unizonal ones. They are 

 widely different, in construction and formation, from the unizonal 

 genicula. 



The number of zones which share the formation of a ffeniculuni 

 in this category, varies according to the species and the individual. 

 In Amp. dilatata the writer counted Io zones (fig. 5) and in Amp. 

 ephedrœa only 4 (fig. 6). In every case there is more than one zone. 

 Solms-Laubach 1j remarked that the genicula of Gorallina are built up 

 with single layer of periclinal cells but those of Amphiroa always two 

 layers. The present writer met with a number of instances which 

 disprove the latter part of his remark. 



]). Corallina p. 29. 



