PAMMEL — POLLINATION OF PHLOMIS TUBEROSA. 257 



hasty in their visits, and therefore always leave some nectar, 

 as can be seen in many cases where the perforator first robs 

 the flowers, after which numerous other insects use the perfora- 

 tions to get nectar, while others again visit the flowers in a 

 normal way. On the whole, I am inclined to think that sterility 

 results more from the disharmonic action of perforating insects 

 than is usually supposed. 



The number of individual flowers perforated by Bombus mas- 

 tivcatus (876) is somewhat startling even in such species as Vicia 

 cracca (87 />, 90), which it could pollinate with perfect ease, yet 

 nine-tenths of all the flowers were perforated. On the loth of 

 June, out of 50 flowers of Gentiana acaulis (90 b) 45 were perfo- 

 rated. Six days later, out of 100 flowers 19 were not perforated ; 

 9 with one, 23 with two, 18 with three, 10 with four, 21 with five 

 openings, so that there were 254 perforations in 100 flowers. 

 Of 50 flowers of Gentiana verna (90 c), 38 were perforated. 



Miiller believes that the work of this insect has caused certain 

 species to become extinct or at least diminished in numbers. 



In England Dr. Wm. Ogle (97 a) found that out of lOD flowers 

 of Melampyrum pratense 96 were perforated. Many observers 

 have noticed that in Trifolhun, Erica, Pentstemon, and other spe- 

 cies, the perforations are quite as numerous. 



Insects certainly show considerable intelligence in making per- 

 forations, or using those made by other insects. One of the most 

 remarkable cases is that observed by Francis Dai'win (28) in a 

 cultivated variety of the Everlasting Pea {Lathyrus sylvestris), 

 where the nectar is enclosed within a tube formed by the united 

 stamens, at the base of which are two natural openings, one on 

 each side, the left being a little the larger. Humble-bees which 

 bit holes through the standard petal always operated on the left- 

 hand side, so as to reach the larger passage. 



The 5 sepals of Aconitum lycoctonum are petal-like and irregu- 

 lar, with the upper one hooded or helmet-shaped: two petals 

 (the other three being rudimentary or absent) consist of small 

 spur-shaped bodies raised on long claws, and are concealed un- 

 der this helmet. Aurivillius (2) has found that in Sweden there 

 are two forms of this flower, in one of which the spur is much 

 shorter than in the other (PI. VII., Fig. 4). Bombus terrestris 

 goes to these short-spurred flowers in a normal way, but it regu- 

 V. — I — 17 



