SECT. IV. OBJECTIONS. 335 



the Ocymum basilicum^ an hermaphrodite plant ; 

 the anthers of several flowers being all cut off 

 before the pollen was ripe, and the stigmas care- 

 fully secured from the access of the pollen of other 

 flowers : in which case it was found that most of 

 the seeds produced were evidently imperfect; 

 though there were also a few that seemed to be 

 completely matured, by their exhibiting on dis- 

 section the same appearances as others that had 

 been exposed to the action of the pollen. But 

 when these apparently perfect seeds were put to the 

 proper test, they were found to be in reality imper- 

 fect ; they did not germinate when sown.* 



This result was discouraging enough, but was not 

 sufficient to deter the Signior from further ex- 

 periment, who now directed his attention to plants 

 of the class Moncecia, to try whether he could not 

 obtain from them a result more favourable to his 

 hypothesis. The subject of his experiment was the 

 Cucurbit a Citrullus, the male flowers of which were 

 destroyed as soon as they made their appearance ; 

 and the female flowers, in order to prevent all 

 suspicion of the access of pollen, were inclosed in 

 bottles luted to the stem by the neck, so as to 

 exclude even the external air. The seeds which 

 were procured in this way germinated and produced 

 plants. 



This result was as favourable to his hypothesis as 

 could be wished ; but to give to the argument 

 * Dissert, vol. iii. chap. i. Eng, Trims. 



