318 PROCEEDINGS OP THE ACADEMY OP [1886. 



fertilization could only occur through some of this extracted 

 pollen escaping from the insect to the stigma. 



However, the fact was clearly demonstrated that Cassia Mari- 

 landica, in this experiment, does not produce a single seed, when 

 the flowers are protected from the visits of insects. 



This plant is the more interesting as it belongs to an order 

 -which the enthusiasts to whom I have referred, see " from its struc- 

 ture " to be so well " arranged for cross-fertilization,'' but which 

 those who, like the following author, have followed results, see 

 just the reverse. " To the casual observer of tj^pical structures 

 the papilionaceous plants must present the most difficult nut to 

 crack for a student of cross-breeding. * * * As might be expec- 

 ted from the structure of the flowers, we have in these plants 

 persistent examples of self-fertilization, and hence the constancy 

 of garden varieties " (Gard. Mag., Feb. 3, ISTT). 



