C. Pkllkw and F. y\. DiKiiAiM 163 



Pollen. 



Darwin was the first to show that there was a structural difference 

 between the pollen grains of the two types in dimorphic primulas, 

 those of the "thrum" plants having a larger diameter than those of 

 " pin " plants. 



In P. verticillatii, where only one form e.xists, we. have not seen 

 differences in size between pollens. The pollen of Kewensis is very 

 irregular in size, and a large number of grains are bad. Scott' regards 

 P. verticillata as being functionally dimorphic, because he found that 

 on self-fertilization often no seed was obtained ; but when fertilized by 

 pollen from another plant the percentage of seed was greatly increased. 

 We have no evidence on this point as we have only self-fertilized 

 verticillata. 



Description of Types. 



In 1911 we began a series of experiments in order to investigate 

 ;uid test the interesting results already obtained at Kew and elsewhere. 

 By the kindness of the authorities at Kew, and Messrs Veitch, we were 

 supplied with abundant material for the purpose. 



P.floribunda. PI. XXV, fig. 1. Whole plant covered with glandular 

 jointed hairs ; these are of two kinds, long many celled pointed hairs and 

 short glandular hairs. Leaves 3 — 6 inches long, ovate, broadly toothed, 

 petiole broad. Scapes numerous, 4 — 8 inches high, bearing super- 

 posed whorls of 3 — 6 flowers, subtended by an involucre of 8 or 4 

 bracts. Corolla golden yellow, limb flat, obcordate. Scentless. Hetero- 

 styled : in the long styled form the stigma is at the mouth of the 

 corolla tube, and the stamens are near the base of the tube : the 

 " thrum " form has a long style, with the stigma and the stamens at 

 the mouth of the tube. There are slight variations in these relative 

 positions Pollen of " thrum " 5 /a, of " pins " 4 ^ in diameter. Habitat 

 Himalayas; figured B.M. t. 6172, vol. 39, 3rd series. Sir George Watt-' 

 gives a description of P. floribunda in its native habitat : " P. flor. occurs 

 in clefts on damp rocks from Kumaon to Simla and Kashmir at altitudes 

 from 3500—6000 ft. In its area however the altitude is gradually 

 lowered on passing westward until in the N. Punjab it occurs almost at 

 the level of the plains, and, what is perhaps more significant still, with 



' J. Scott, Journ. Linn. Soc. Vol. viii. 186.5. 



- Sir George Watt, J. R. H. S., "Report of Primula Coufereuce," Aug. 191.5, p. I'.tl'i, 

 reprinted from J. R. H. S., 1904, p. 29-5. 



U—-i 



