STUDIES ON -ONONIS IN BRITAIN 



313 



Analysis of Individual Chaiiacters. 



The way in which intermediate characters tend to occur in the same plants is not 

 displayed to full advantage by the pictorial scatter diagrams, owing to the unreliability 

 of leaf measurements as an indication of hybridity. These were the only quantitative 

 characters, amongst those scored, which were sufficiently variable to serve, as the axes in 

 the diagrams. A clearer indication of this tendency for intermediate characters to be 

 associated with each other can be obtained from an analysis of the individual characters 

 which were used in compiling the character sum index, and also from the additional 



Fig. 7. Mature legume and calyx of O. repens (left) and O. spinosa (right). X 2. 



character of fruit index*. The same pure populations were again used for purposes of 

 comparison. The results for each character, in the hybrid populations, were separated 

 into three groups on the basis of the character sum index, (a) plants scoring under 13 

 (i.e. resembling O. repens), (b) plants scoring 13 to 17 (i.e. intermediate), and (c) plants 

 scoring 18 and over (i.e. resembling O. spinosa). The results for leaflet and fruit index 

 are expressed in Fig. 8, and for the remaining characters, with the exception of flower 

 indexf, in Table 2. They show an obvious correlation between each character and the 

 character sum index. In all cases (except in the spinous condition at Marsden) intermediate 

 characters are predominantly associated with intermediate plants, and are not distributed 

 throughout most of the individuals in the colony. The occurrence of non-spiny plants 

 of O. spinosa (var. mitis) at Marsden is shown in Table 2 in which 22-5% of the plants 

 with a character sum index of 18 and over have no spines. This condition is not entirely 

 due to hybridity, for similar plants have been noted in apparently pure populations of 

 this species (e.g. in the Quarrington area - cf. Table 2 in which 2% of the plants which 

 were scored had no spines). 



Conclusions 



Two conclusions may be drawn from these results; one relating to the method 

 used in analysing the colonies, and the other to the colonies themselves. In so far as the 

 method is concerned, the results obtained from the analysis of the pure colonies give a 

 clear indication that the characters, when used together, provide a reliable means of 

 separating these two species. The value of the individual characters varies considerably 

 and it is probable that no single character would give a satisfactory separation. Separation 

 of the two species should be equally effective in a mixed colony provided that the species 



• A fruit index was obtained by dividing the calyx length by the legume length (see Fig. 7). As mature fruits cannot always be 

 obtained at the same time as well-developed flowers, this character could not be included in the character sum index, which was 

 prepared at the time of flowering. For this reason the analysis of this character was based on the appearance of the plants, whether 

 identifiable as O. repens, O. spinosa, or intermediate, and not on the character sum index. 



t Flower index was omitted because, on analysis, it was found that the relative lengths of the wing and keel showed no significant 

 differences between the two species, in the colonies which were studied. It would therefore appear probable that this character is 

 of no value in separating these two species, though it is sometimes used in floras. 



