64 MORPHOLOGY OP ANGIOSPKRMS 



scantia virginica as having one. Among Liliaceae, Allium, 

 Hemerocallis, Lilium, Erythronium, and Tricyrtis have no pa- 

 rietal cell; and Oonvallaria, FunJcia, S cilia, Ornithogalum, Tril- 

 lium, and Yucca are known to have one. Among Iridaceae, the 

 only records we have been able to find are those of Sisyrinchium 

 iritli folium (Strasburger 1S ) and Iris stylo sa (Gnignard 17 ), in 

 neither of which is there a parietal cell ; but it would be very 

 unsafe to predicate this condition for the whole family. Among 

 the Cannaceae, Gnignard 17 reports Carina indica as sometimes 

 having a parietal cell and sometimes not, but Wiegand 54 finds 

 in it only an abundant parietal tissue; and the other Scitami- 

 neae are reported by Humphrey 32 with parietal tissue. Among 

 Orchidaceae, Gymnadenia conopsea (Strasburger 13 ) and Or- 

 chis pollens (Goebel, 20 p. .')'.) 1) were long ago reported as with- 

 out a parietal cell, but recently Dumee, 44 examining a number 

 of genera and species of orchids, reports them all as having 

 parietal cells. This record probably fairly represents the con- 

 dition of the parietal tissue among Monocotyledons. It indi- 

 cates a general tendency to suppress it, which has been success- 

 ful in certain members of the higher and more specialized 

 families. 



Among the Archichlamydeae approximately the same con- 

 dition prevails. The Ranunculaceae exhibit a surprisinglv uni- 

 form suppression of the parietal tissue, this condition having 

 been found in Anemone, Caltha, Clematis, Delphinium, Myo- 

 surus, and Ranunculus (Fig. 27) ; while in Aquilegia a parietal 

 cell may or may not appear. Only Helleborus (Gnignard 17 ) 

 and Thalictrum (Overton 59 ) have thus far been reported as 

 having a parietal cell, and this may or may not divide. It is 

 to lie noted that in Delphinium, Caltha, and Jeffcrsonia the 

 absence of parietal tissue is compensated tor by numerous peri- 

 clinal and anticlinal divisions of the overlying epidermal cells; 

 and in the Balanophoraceae this epidermal growth reaches so 

 remarkable a development that Treub at first called it a style. 

 The same development is seen in Hippuris (Fischer 1B ), in which 

 the apical epidermal cell divides by anticlinal and periclinal 

 walls and forms a small, wedge-shaped cushion that prevents 

 the micropyle from being entirely obliterated by the closing 

 in of the integument. Among the Berberidaceae, Jeffersonia 

 (Andrews 31 ) has no parietal cell, and Mahonia indica (Guig- 



