45 



extend all around the canal, but often covers only patches of its 

 wall; it frequently passes over some of the hairs, covering and 

 enclosing them entirely. In cross-sections the outer edge of the 

 lining IS firm and smooth, assuming a distinct yellow color in 

 chloriodide of zinc, or in potassic iodide and sulphuric acid: it 

 is, in fact, a thin layer of cuticle raised in a coherent film from 

 the wall of the canal by the mucilage forming under it Wher- 

 ever the mucilage occurs calcium oxalate is nearly always found 

 in abundance ; not, however, inside the cells or in separate recep- 

 tacles, but imbedded in the mass of the mucilage in the form of 

 crystals, usually octahedral, which are frequently of extraordinary 

 size and beauty. Quite often large crystalline conglomerates 

 cling to the sides of the canals, which are always coated besides 

 with a dense crystalline layer of this salt. 



The internal hairs have a thin membrane, and most of them 

 bear on their surface numerous bubble-like excrescences, which 

 sometimes attain considerable size (Fig. 6). Some rest on the 

 hair on a broad circular base, others barely touch it at one point 

 (Fig. 6). In the former case, when their elevation is very slight, 

 an exceedingly thin membrane, the continuation of the outer- 

 most layer of the wall of the hair, seems to cover them. This may 

 be noticed after treatment with nitric acid and subsequent appli- 

 cation of chloriodide of zinc. I was unable to discover, with any 

 degree of certainty, the least trace of a membrane around the 

 globular bubbles. Other bubbles of exactly the same appearance 

 and structure are frequently found on the sides of the canals, cling- 

 ing to the walls of cells that have not grown into hairs. 



Mellink* noticed ''small hemispherical,! or more rarely pedi- 

 celed bubble-like elevations" on the hairs of NymphcBa, and 



L 



reaches the conclusion that the cuticle of the hair is thickened in 

 some places so as to cause them. The bubbles on the hairs of 

 Brasenia are not stained by chloriodide of zinc, potassic iodide, 

 or nitric acid | (while the cuticle of the epidermis and the suber- 

 ized parts of the endodermis react beautifully); they are evidently 



t " rialbkreisformige," which I suppose ought to read " halbkugelformige." 



X Mellink does not report the application of reagents to prove that the bubbles are 



really a thickening of the cuticle, but simply says that they are not hollow, because 



acetic acid-rosanilin will stain them uniformly red. 



