NUPHAR INTERMEDIA IN BRITAIN 



19 



data obtained from them are compared with equivalent data derived from 105 specimens 

 of British N. lutea and 58 specimens of British N. pumila. Naturally not all of the herbarium 

 specimens have been equally favourable for the assessment of the selected attributes, but 

 in no case do the means of Table 4 refer to less than 15 individuals for N. intermedia, 16 

 for N. pumila and 51 for N. lutea. 



Direct comparison between the data for Table 4 and those of Table 1 is not per- 

 missible except in the case of carpel number, because of the considerable shrinkage 

 resultant from drying in herbarium specimens, which ranges from 20-30 per cent. 

 However, it is evident that the size relationships revealed by the herbarium samples of 

 the three taxa are closely similar to those indicated by the fresh samples, the proportional 

 differences in sepal size being almost identical (see Fig. 12). The same is true also of 

 peduncle and carpel diameter. The correspondence of carpel number (a feature not, of 

 course, affected by drying) between the herbarium samples and the fresh ones is quite 

 remarkably close for N. lutea and N. pumila (Fig. 1-1). The difference between the means 

 for carpel number of the herbarium and fresh samples of N. intermedia (1-16) is quite 

 small, although just significant statistically (p = -04). 



I , 1 , 1 1 i 1 — I 



9 10 11 12 13 14 15 16 



Carpel number 



Fig. 11. The relation of carpel number to carpel diameter in European Nuphar taxa. Series 1 from living 

 British plants; Series 2 from British plants ex herhariis; and Series 3 from continental plants ex herhariis. 

 P = N. pumila; I = N. intermedia; and L = N. lutea. The means of each sample are inserted, and distances 

 corresponding to twice the standard error are indicated either side of the mean. The Avinlochan colony 

 (" introgressed " N. pumila) is indicated separately. 



N. intermedia in Continental Europe 



As with material from British sources, the specimens of the three Nuphar taxa 

 from continental localities in the British Museum and Kew herbaria have been employed 

 as samples for the purpose of a biometrical study. The numbers of plants examined 

 were 99 for N. lutea, 92 for N. pumila and 87 for N. intermedia. Data from these samples 

 comparable with those of Table 4 for British samples are given in Table 5. The same 

 size relationship prevails between the three taxa as in the British fresh and British her- 

 barium material (Figs. 11 and 12). The continental and British herbarium samples of 

 N. pumila and N. intermedia are closely similar in size, carpel number, ovary diameter 

 and peduncle diameter, none of the differences being statistically significant. In N. 

 lutea, the mean values of sepal length and width for the continental herbarium sample 

 are slightly smaller than for the British herbarium sample. The carpel numbers are 



