1920] SNOW— DIAPHRAGMS 305 



of the 29 cases of a change in rate of growth and diaphragm distance 

 shown in table IV, 19 (64 per cent) showed an inverse relation 

 between the two; 9 (24 per cent) showed a direct relation; and in 

 3 cases (10 per cent) an equal distance went with an increased rate. 

 When we remember that the normal tendency is to increase the 

 distance from tip to base, these last three cases really show an 

 inverse relation which, added to the 19 preceding, make 22 (75 per 

 cent) which show an inverse relation. Of these, 14 (49 per cent 

 of total) show an increase downward, which coincides with the 

 normal tendency, while 8 (27 per cent of total) show a decrease 

 downward in opposition to the normal tendency, and therefore 

 the more significant. These confirm the indication shown by the 

 49 per cent, and tend to establish an inverse relation between rate 

 of growth and distance between diaphragms. 



This is not what one would expect if the distance between dia- 

 phragms is considered to be brought about by the excessive growth, 

 or stretching, of the intervening tissues. It is what one would 

 expect, however, if, as was suggested in a former paper (13), dia- 

 phragms are due to certain cells retaining their power of division 

 and growth, while those above and below them lose this power 

 and are drawn out into arachnoid cells by the growth of the sur- 

 rounding tissues; and also if the formative activities show a 

 gradient from beginning to end of the growth of the stem. This 

 suggests several interesting questions. Is there such a formative 

 gradient? Would the respiration test show a gradual decline in 

 metabolic changes, or would it follow the growth curve? Is it 

 possible that, in averaging the diaphragm distances in a region, a 

 shortening, corresponding to the rise in the growth curve, was over- 

 looked ? Is the peak of the growth curve due wholly to a stretching 

 period ? If so, would this stretching counterbalance the tendency 

 to shorten the diaphragm distance with the rise in the growth curve ? 



Air chambers. — Scirpus validus appears to start with four 

 large chambers and a number of small ones. As the culm grows, 

 the small ones increase in size, until many nearly equal-sized 

 spaces are the result. The different culms may pass through these 

 stages at different rates; therefore the same regions cannot be 

 compared. Only two cases were noted in which a rather rapid 



