456 BOTANICAL GAZETTE [juxe 



it does not appear to be very extensively sought by insects, but species 

 of Syrphidae may be seen flitting from flower to flower of this and 

 of plants of Krigia virginica in blossom at the same time, perhaps 

 as much in search of pollen as for their sweets. 



Muller examined other European species of Linaria with refer- 

 ence to this matter, among them L. minor Desf. and L. arvensis L. 

 Their flowers are very small, but adapted, like L. vulgaris and L. 

 alpina, to pollination by bees. As a weed in his garden at Lipp- 

 stadt he "looked in vain" for visitors to L. minor, and L. arvensis 

 was repeatedly watched in favorable weather in another station with 

 a like result. Hence he concludes that they are restricted to self- 

 fertilization. As the anthers burst at the same time the stigma 

 matures, should a bee come for the nectar the flowers are ready for 

 cross-pollination. This condition lasts only a short time; the stigma 

 is soon covered with pollen, and self-fertilization is accomplished. 

 Since Muller cannot imagine that a flower, in all the peculiarities 

 of its structure fitted for pollination by insects, should still be very 

 exceptionally visited and crossed by their instrumentality, he con- 

 cludes that we have in these plants a deteriorating descendant of an 

 ancestor with larger and more striking flowers, in whose pollination 

 bees as a rule took part. 6 He considers that the 

 various other plants with diminutive or inconspicuous mellitophilous 



same 



ery rarely visited by bees, citing amon 



Vicia hirsute Koch as a similar case, whose style bears unequivocal 



marks 



hairs at most. 7 



6 



II. THE CLEISTOGAMIC CONDITION 



Muller 



mention 



that they do not open at all, or the cleistogamic stage, in which self- 

 pollination is the only means of securing fertility. Of the eight 

 types of entomophilous flowers made by Delpino, the sixth is that 

 in which the anthers and stigmas are close together and included. 



linaria answers these conditions, as must indeed be the case with 



6 Muller, Weitere Beobachtungen III. Verh. nat. Vereins Rheinl. Westf. 39 : * 8 - 

 1880. 



, Ibid. II. Op. cit. 38:360. 1879. 



1 



