External Conditions and Manurinij. 319 



I counted the sepals of all the flowers. There were about 

 2500 on the manured bed, and about 1500 on the un- 

 manured. Amongst them were many with five and f(jur 

 sepals, and about 20 with 3, but none with fewer tiian 3 

 or more than 5. Here again, a pronounced half curve 

 was the result. I have reckoned together the i)roportion 

 of 3-4-merous flowers for the individual counts, and at 

 each stage in the counting collected all the flowers wliich 

 had opened since the preceding one. The counts were 

 made when possible every fourth day, or, when the num- 

 ber of flowers was too small, at greater intervals. The 

 result was as follows : 



PROPORTION OF 3-4-MEROUS FLOWERS IN %. 



June July Augus. 



Day: 19 23 27 15 9 13 17 21 25 29 2 



Manured: 7 13 24 28 34 39 50 65 49 49 43 27 % 



Unmanured: — — 7 — 20 33 39 — 42 49 46 44 % 



We see that the proportion of anomalies increased 

 on both beds gradually throughout the summer, reached 

 its maximum in the second half of July, and then sank 

 again. On the manured bed, however, this proportion 

 amounted to 65% of the whole and on the unmanured 

 bed to 49% of the flowers counted (160 and 224 flowers 

 in the two cases respectively). In this case, therefore, 

 both the periodicity and the relation to the external con- 

 ditions are in all essentials the same as in the case of the 

 wliite clover. In this latter case a plant which T had 

 raised from seed served as material; but in tliat of 

 PotcntiUa a specimen which T liad collected in tlie fleld. 



I shall now deal as briefly as possible with a series 

 of further instances, emphasizing as before that the ex- 

 ternal influences have the results in question only when 

 the particular cliaracters are already present in the semi- 



