NOTES ON POLLINATION. 19 



species liave the stigmas so placed as to be certain of contact with 

 an entering insect. The stigma in M. pilosus occupies nearly 

 nil of the narrovt- throat. The throat of 31. moschatus is much 

 narrowed by two ridges, and the stigma lies below this narrowed 

 entrance. The palate of M. liiteus does not quite close the throat, 

 and small flies, beetles, etc., sometimes enter without striking the 

 stigma, but tlie dense hairiness of the lower lip makes this occur- 

 rence rare, and a large bee is almost sure to eftect cross-pollination. 

 M. prlmuloldes has a more open throat, and it is only in the opening 

 bud that entrance to the honey is fully guarded by the stigma; I 

 found a considerable number of stigmas of this species insensible to 

 irritation, and inferred that they had been pollinated in the bud. 

 M. Falmerl, like M. cardinalis, holds its essential ox'gans a consider- 

 able distance above the open throat, but I do not know whether this 

 red flower can, like its very showy relative, depend on humming- 

 bird visits. I saw none of these species frequently visited. 3f. hiteus, 

 which I have never seen visited by insects in lower altitudes, not 

 rarely in the mountains, entertains Bomhiis Galifornieus and some- 

 times smaller insects. I found thrips in M. moschatus, and found 

 flowers with stigmas visibly pollinated before the deliiscence of their 

 own anthers. I saw M. primuloides only casually visited by bees. 

 This species has, late in the season, many small flowers only three or 

 four lines long, with styles extremely variable in length. Sometimes 

 the styles are so short as to bring the stigmas below the anthers, one 

 stigma lobe being very small in these cases, and so self-pollination is 

 insured. I found but one similar instance in the other species, this 

 being a small late flower of Iil. moschatits. 



PentstemOxV Palmeri, Gray, P. barbatus, var. labrosus, Gray, 

 and F. Bridgesii, Gray. These Pentstemons all have abundant honey 

 secreted by the bases of the two upper perfect stamens. The fila- 

 ments of these stamens curve inward to meet each other and tlie 

 other two anther-bearing filaments, then all four filaments sweep 

 over so as to lie against the upper Avail of the flower. The sterile 

 filament crosses the tube above the junction of the others and lies 

 for the rest of its length against the lower wall; all of which 

 mechaiiism has the eflect of excluding short-tongued guests from the 

 honey. The anthers do not dehisce simultaneously, and the process 



