STELLARIA NEMORUM L. SUBSPECIES GLOCHIDISPERMA 123 



The edge of the ripe seed of subsp. nemorum is furnished with rows of rounded 

 more or less hemispherical tubercles, whilst the margin of the seed of subsp. glochidisperma 

 has long cylindrical papillae about 0-15 mm. in length, which under the microscope 

 are seen to possess a barbate cap. A warning should be added that when the seed is 

 immature the marginal tubercles in subsp. nemorum may seem longer in proportion to 

 breadth than when mature ; especially is this the case in herbarium material where the 

 immature seed has become somewhat shrivelled. 



A further characteristic, although not diagnostic in itself, is that in subsp. glochidi- 

 sperma the uppermost leaves are stalked whilst those of subsp. nemorum are usually sessile 

 or subsessile. This is an unreliable character where the specimen is either a very lush 

 nemorum or a diminutive glochidisperma, whilst in young growth and on vegetative shoots 

 the leaves of both subspecies are stalked. Typically, a flowering specimen of subsp. 

 glochidisperma has the appearance of Circaea lutetiana; a figure is to be found in Hegi 

 (1911 a or b) under the synonym, subsp. circaeoides A. Schwarz. 



In addition to these distinguishing characters, Murbeck in his original description 

 of subsp. glochidisperma (Murbeck 1892) mentioned several others which on being tested 

 have proved to be either of little value or else to be common to both subspecies. (1) It was 

 claimed that the stem leaves of subsp. glochidisperma are broader. If breadth alone is 

 considered this is quite false, and even if breadth is considered relative to length there is 

 no difference between the two subspecies in ordinary vegetative leaves. Only with the 

 pair of leaves at the last node before the first branching of the cyme does there seem to be 

 any difference, and even then it can only be said that there is a tendency for subsp. 

 glochidisperma to have leaves broader in proportion to their length. In the specimens 

 measured, the index of leaf-breadth divided by leaf-length for subsp. nemorum was 

 0-31-0-59 and for subsp. glochidisperma 0-39-0-70. (2) Subsp. glochidisperma is said to 

 possess more cordate leaf bases. Examination again shows that this is not so in purely 

 vegetative leaves and that only in the uppermost ones, including the first pair of bracts, 

 is there a slight difference. In subsp. nemorum they are truncate to rounded, whilst in 

 subsp. glochidisperma they are subcordate to truncate. Furthermore, there is a tendency 

 in both subspecies for the leaf blade to be slightly decurrent. This tendency is greater 

 subsp. nemorum, partly because the leaf base is less cordate and partly because the 

 petiole is usually shorter. In leaves which are very shortly stalked or subsessile the d ecurrent 

 wing makes the leaf base merge into the axis of the shoot. (3) The capsule is said to be 

 twice as long as the calyx in subsp. glochidisperma and only a third longer in subsp. 

 nemorum. This has proved a difficult character to test, as relatively little of the available 

 material has fully ripe capsules. It seems there may be a tendency in subsp. glochidisperma 

 for the capsules to be longer in proportion to the calyx length, but it is doubtful if it is as 

 significant as has been claimed by Murbeck. (4) Another character that has not been 

 tested is that the fruiting pedicel in subsp. glochidisperma lies in line with the axis whilst 

 in subsp. nemorum it lies bent at an angle. It is quite impossible to observe this with 

 certainty on herbarium material and field or experimental garden studies will be necessary. 



Early in this investigation, before it was known that the chromosome numbers of both 

 subspecies had been found to be the same (2n = 26, Peterson 1936), it was thought possible 

 that one of them might be a polyploid. Had this been so it might have been reflected 

 in the pollen grain sizes and stomatal indices of the two subspecies. No significant difference 

 in pollen grain size between the two subspecies was found, however. The pollen of six 

 specimens of subsp. glochidisperma was measured and the total range in diameter found 

 to be (40 — ) 41-48 (— 50) /x, whilst in nine specimens of subsp. nemorum the total range 

 was (39 — ) 41-47 (— 50) fx ; ten grains per specimen were measured. No comparison 

 has so far been possible between the stomatal indices, that is in the proportion of stomatal 



