288 MacMillan: Oi^servatioxs on Nereocystis 



develop the thcrmometer-tubc cell with capillary lumen such as 

 characterizes the sieve-tube. Since the morphological position of 

 the two kinds of cells in the stipe is different, since their origin is 

 different, and on account of their very easily distinguishable struc- 

 ture, I can but follow Oliver and maintain them as separate 

 structural elements not to be confused on account of their similar 

 perforated end-plates. They are abundant in older material but 

 the relative proportion of the two varieties of cells changes. 

 While in young material an abundance of sieve-tubes is character- 

 istic, in old material the trumpet-hyphae are abundant and but few 

 sieve-tubes remain, most of them having undergone extreme at- 

 ten nation and subsequent degeneration. 



Structure of nuiture Stipe,-. — Cross sections through the stipe of 

 a plant 25 meters in length showed its diameter to be 10 mm., of 

 which the central cylinder comprised 2.5 mm. In this material 

 the cambium zone lying four or five layers within the epidermis is 

 still distinct. Numerous mucilage canals are present, most of them 

 lying in the concentric circle first developed and now separated 

 from the periphery of the stipe by numerous layers of cortex 

 developed after their formation from the cambium. The average 

 size of cortical cells in cross section is about 25 mic. Their walls, 

 however, are now distinctly thicker than at first and at the angles 

 between the cells the thickening is sometimes increased giving to 

 the tissue a collenchymatous appearance. By the thickening of 

 the walls great elasticity and strengtli is given to the stipe. Inter- 

 mingled with the large cortical cells are many smaller ones averag- 

 ing 10 mic. in diameter. The central cylinder in the mature stipe 

 consists of a loose pith-web of anastomosing filaments, imbedded 

 in gelatine, among which abundant large trumpet-hyphae are ap- 

 parent. ]\Iany of the trumpet-hyphae connect by lateral proc- 

 esses with the ordinary filaments which are not more than half as 

 great in diameter. In material of this age the sieve-tubes are very 

 difficult to find, almost all of them having been destroyed. This is 



L 



not what Oliver means by his phrase ** obliteration of sieve-tubes" 

 for he applied that term to the closing of the plates by his callus- 

 like substance, nor can I learn that the fact has previously been 

 recorded. Oliver mentions that the true sieve-tubes are abund- • 

 antly branched. In my opinion this is a mistake. Anastomoses 



