4 Transactions of the Society. 



mainly, if not entirely, upon increase in size of the cells without 

 cell division, measurements were made of the dimensions in trans- 

 verse section of ten pith cells taken at random from the extreme 

 bases of the inflorescences gathered on February 9 and April 27 

 respectively. The pith cells in the former case were found to- 

 average 58 /a and in the latter 73 /a in diameter, which represents 

 an increase of 26 p.c. The entire axis had increased in the same 

 period from approximately 6 mm. to 7*5 mm. in diameter, repre- 

 senting an increase of 25 p.c. As there is much variation in size 

 in the pith cells, no great reliance can be placed upon the exact 

 figure obtained for their average diameter, and the extremely close 

 coincidence of these percentages is probably more or less accidental y. 

 but we may at least conclude that in the case of the pith the 

 stretching of already existing cells is competent to account for the 

 increase of diameter, and that no appreciable amount of cell mul- 

 tiplication, with formation of cell walls in planes parallel to the 

 long axis of the organ, need be postulated. 



In describing the binucleate phase in the inflorescence gathered 

 on April 27, we have so far been considering only the main part of 

 the flowering region. Binucleate cells occur near the apex, but 

 this is not a favourable case for studying the exact point at which 

 this condition arises, since the tip becomes hollow and dies at 

 a relatively early period. Passing to the lower sterile region 

 of the inflorescence axis, we find that binucleate cells no longer 

 occur, but the single nuclei are bilobed or irregularly lobed. 



For comparison, older and younger inflorescences were studied, 

 A very young inflorescence was dissected out of the terminal leaf- 

 bud early in February. The flowering region was about 2 • 5 cm, 

 long, and the sterile stalk not more than 2 cm. Binucleate cells 

 and phragmospheres occurred in the fertile region, just as in that 

 gathered at the end of April. In this young inflorescence, how- 

 ever, phragmospheres were not confined to the fertile region, but 

 occurred, in addition, in the sterile stalk. 



Two older inflorescences, gathered on May 7 and May 28 re- 

 spectively, were also studied. On examining the fertile region of 

 the inflorescence fixed on May 7, it was found that, though many 

 cells of the pith and ground tissue were binucleate, as in the 

 younger axes, nuclear division had apparently ceased and many of 

 the cells had become uninucleate. The nuclei within the uni- 

 nucleate cells were often neatly bilobed, but sometimes lobed o? 

 fissured in an irregular fashion, thus recalling the condition in the 

 sterile lower region in the inflorescence axis fixed ten days earlier 

 (PI. I, fig. 10). In the axis gathered on May 28 uninucleate cells 

 liad become still more universal, and the lobing was as pronounced 

 as in the previous case. In both these axes the nuclei appeared 

 somewhat flattened wlien seen in profile. The bilobing of the 

 nuclei was so frequent and striking that its occurrence in cella 



