254 On the Cross-Fertilization of Flowers by Insects. 



cross-fertilization " *. It is peculiarly characteristic of the 

 present clay to seek for the effect of external conditions and to 

 experiment with the modifications that can be brought about 

 by changing these. For example, Prof. Goebel says, in 

 reference to cleistogamic flowers, " We do not yet know the 

 conditions necessary for the production of cleistogamic flowers, 

 but it may be assumed even now that this production is 

 influenced by external factors wherever a plant has the power 

 to produce such flowers " -f\ He then cites experiments made 

 with Impatiens fulva, where cleistogamic flowers are the 

 result of poor nutrition. Accordingly it appears at least 

 possible that experiments might give similar results in 

 reference to dichogamy. 



Attention has often been called to the fact that in plants 

 especially adapted to insure cross-fertilization there exists in 

 almost every case a possibility of self-fertilization. The above 

 statements in regard to dichogamy, if of any value, point to 

 a different conclusion concerning the final or at least possible 

 autogamy from that which is generally accepted. It is 

 assumed that the flower is so constructed that there is every 

 chance of a cross provided the insect appears to do the work. 

 This failing, the arrangement is such as to allow pollen to 

 come in contact with the stigma of the same flower. It is 

 evident that in every case which seems to point towards cross- 

 fertilization it is always to a great extent a matter of chance 

 whether the visitor arrives or not, even when the adaptation 

 seems most perfect. It should be decided if cross-fertilization 

 or autogamy is the rule with every species which seems con- 

 structed so as to attract insects, and this work should be clone 

 in as many different localities and at different times of the 

 year as possible, since there is no doubt there are great 

 variations possible in the fertilization of flowers in the same 

 species caused by different conditions of heat, moisture, &c. 



If autogamy should in any case prove the rule, we must 

 regard fertilization by aid of an insect as an exception, not to 

 call it an accident. Dichogamy probably is then in a measure 

 due to external conditions. If this is true, it is simply the 

 result of a " lagging behind " in the ripening time of either 

 stamens or pistils, and the final autogamy is the result of a 

 subsequent " catching-up " in this respect. This is, as I 

 have said before, probably the most interesting side of the 

 question, and the one which will no doubt prove the most 

 satisfactory for experimental investigation. 



* Proc. Acad. Nat. Sci. Pbilad. Nov. 27, 1888, p. 394. 

 t Goebel, ' Pflauzeu-biologisclie Sckilderungeu,' Marburg, 1893, ii. Tkeil, 

 2 Lief. p. 363. 



