F. H. WHITEHEAD 



219 



it is thought that the points raised in its consideration are of such importance that they 

 should be mentioned. 



An examination of many plants showed that there was considerable variation in the 

 angle formed by the outer branches of the dichasium. Correlated with this character 

 was a variation in the angle made with the ground by the vegetative branches and the 

 lower parts (i.e. below the first dichasium) of the flowering branches. 



Investigation showed that these two sets of characters can be taken as expressing 

 the characteristic physiognomy of the plant. It was found that there appear to be differ- 

 ences between the means of the primary groups for the character dichasial angle. 



Observations made in the field, however, gave the impression that the plants with 

 large dichasial angles were well-grown plants growing apart from other plants. On the 

 other hand, plants with small dichasial angles were always growing amongst longish 

 grass or closely surrounded with other Cerastium plants. It appeared from these obser- 

 vations that the dichasial angle might be varying more or less completely in response to 

 environment and its use as a taxonomic or diagnostic character would be open to con- 

 siderable doubt. Since a number of varietal names have been given to taxa distinguished 

 by this type of character it was considered of importance to ascertain its reliability. 

 A preliminary investigation of the relationship of these characters to environment 

 (Whitehead 1955) showed that they were not suitable for use in discriminating between 

 these groups. 



It remains to account for the observed differences of dichasial angles in the field 

 and herbarium material. If the usual habitats of the three taxa are considered, the observed 

 differences are not so surprising. C. atrovirens is a plant of sand-dunes and cliff turf, 

 usually very short, and is only occasionally found inland, growing among taller vegetation. 

 Tall plants with smaller dichasial angles are often found at the entrances to rabbit burrows 

 but the majority of plants are from habitats fully exposed. Even where there is great 

 mutual competition the plants are usually short and do not shade each other to the extent 

 that they would if their growth were more luxuriant. 



To some extent these remarks apply to C. semidecandrum, but it is more frequently 

 found inland than C. atrovirens, where it can be found, sometimes, as a fairly large plant 

 up to 30 cm. high growing amongst tall vegetation; nevertheless, it is more often found 

 growing on rather exposed patches of soil and forming part of the ephemeral element 

 of vernal annuals. It is also found associated with C. atrovirens on sand-dunes and, as 

 far as physiognomy is concerned, is then identical with it. 



C. pumilum, however, despite its occurrence in grazed turf and amongst limestone 

 detritus in some localities, is much more often found growing among rather taller 

 vegetation. From its original description the impression is obtained that it is a very 

 dwarf Cerastium but an examination of the range of this taxon shows it to be quite as 

 tall as the other taxa and with a greater frequency of taller plants. 



It can be seen that C. pumilum is the taxon most frequently found in competition 

 with the taller vegetation and it was found to have rather narrower dichasial angles than 

 the other taxa, C. atrovirens having the larger dichasial angles and being more restricted 

 to exposed habitats. C. semidecandrum, which has the widest range of dichasial angle, 

 is also the taxon with the greatest range of habitats. 



The Character Calyx Length/Capsule Length 



This character has long been used diagnostically and it was hoped that it might be 

 used in this analysis. One limitation of the statistical treatment which follows is that 

 a value for each character for each specimen is absolutely necessary. It was found that 



