60 THE WATERLILIES. 



the exchange of gases between the air in the canals and the surrounding 

 tissues. For, following Goebel (1893), it seems generally accepted that 

 the air-spaces of aquatic plants are for the purpose of internal breathing ; 

 and the researches of Lechartier (1867) as well as those cited by Goebel 

 (1. c.) are in favor of this view. Such an internal cuticle can hardly be 

 for protection in the same sense as this function is attributed to it on the 

 outer surfaces of plants, though it might be simply a chemical product of 

 the contact of air upon an otherwise unmodified living cell-wall. Similar 

 arguments apply to the cuticular coatings on the surfaces of mesophyll- 

 cells where in contact with air, as in the air-chambers connected with the 

 stomata in most plants. Russow (1884) and Mellinck (1886) regarded 

 these coatings as intercellular protoplasm; but Kny ( 1900), working over 

 the same ground very recently, considers the presence of living extra- 

 cellular protoplasm in the large air-canals of water plants improbable ; 

 the writer favors this view for Nymphaea. Kohl (1889) found on the 

 walls of the air-canals of N. alba small free crystals of calcium oxalate. 



Between two adjacent air-canals there is often but a single layer 

 of cells (e. g., N. flai'a, tetragona, and small leaves of riibra, amazoniun, 

 etc.); usually, however, the partitions are 2 to 6 or even 12 cells thick. 

 The thickness varies slightly in different parts of the same petiole, since 

 the canals taper a little at each end. The largest ones, at least, open at 

 the proximal end into the lacunar mediocortex of the stem by pores in 

 the end-wall ; at the distal end they connect similarly with a mass of 

 similar lacunar tissue which again makes communication with the air- 

 canals of the veins of the leaf. At each of these points of opening the 

 walls of the canals and the lacunar tissues are loaded with stellate idio- 

 blasts ; indeed, the lacunar tissue in N. /lava, odorata and tnberosa seems 

 to be little else than a tangled mass of such cells, with their arms inter- 

 lacing in every direction. Diaphragms do not occur in the air-canals of 

 the petioles and peduncles. 



It will be well here to dispose of some matters relating to the inter- 

 nal hairs of Nymphaea. We have adopted for them the Sachsian term 

 "idioblast" as used by Weiss (I878). 1 Their common features are the 

 great induration of the walls, and the presence of numberless small 



1 The earlier terms " pachycyst " of Caspary (1865), " pneumatocyst " of Planchon (1850, e) 

 and " Knorpelzelle " of Hanstcin, have never come into general use. Solereder's (1898) 

 " Sclerenchymzelle " is a good term, but a little too broad. Ischirch (1889) has proposed the 

 term " astrosclcreide," but Arcangeli (1890, a) rightly objects to this because so many forms are 

 found in addition to the stellate-cell ; he therefore proposes " cladosclereide." 



