550 



Dr. W. L. Balls. The Existence of Daily 



The lint-hairs also show lamination at this stage for the first time, but 

 none have been observed on which dependable counts could be made. 



Upon counting the number of growth-rings in the wall of 24 fuzz-hairs, the 

 following frequency was obtained : — 



These figures are too small for any statistical generalisation, but it is 

 evident that they vary only a little more than we anticipated, and that the 

 mean is approximately three rings higher as the result of three days more 

 growth. 



Thirty -three-day Fruit (fig. 4). — The general appearance of this stage is 

 like that of the previous one, but the count of growth-rings in the fuzz gave 

 a mean value of 5"3, being 2"7 higher than the fruit of three days previously, 

 with the following frequency distribution : — 



Eings 1 23456 789 10 11 



Number of hairs — — — 4433 — 25 1 1 



Photographs of one lint-hair showing ten rings, are reproduced in 

 figs. 6 and 7, Plate 16. 



Thirty -six-day Fruit (fig. 5). — The increase in the number of rings in this 

 glide merely brought the value up to 6'6 on the first count taken, for reasons 

 which will shortly be described. The frequency distribution was : — 



Rings 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 



Number of hairs 0001141455 3 2 



The scatter ^of the observations is much the same as before, since 3 per 

 cent. P.E. would allow seven days extreme variation at this age, against 

 nine days actually observed, but the next two stages show that an obser- 

 vational error is creeping in. 



Meanwhile we may note that four lint-hairs in this stage were in a state 

 which enabled reliable counts of their growth-rings to be made, and were 

 found to contain 7, 10, 11, and 12 rings respectively. 



Thirty-nine-day Fruit. — The counts of the fuzz-rings were substantially 

 identical with those from the previous stage, but the preparation was better 



Rings 



1 2 3 4 5 6 7 



Number of hairs 



02385312 = 24. Mean = 3-6 



