-235— 



which can only be maHe satisfactorily at night with the assistance <»f a 

 confederate. If Mr. Hulst has any valid reasons for not coming to the 

 sa ne conclusions, he will better advance our knowledge by giving his 

 reasons than by the mere expression of an opini(tn without evidence thai 

 he is in any way justified in it. Until he does the presumption is that 

 he has carelessly observed or has even confounded Pronuba and Prodoxus 

 HS others have done before him. 



Finally. I commend to Mr. Hulst a more careful study of my writ- 

 ings upon this subject and especially those in the Proceedings of the 

 .\merican Association for the Advancement of .Science for i88oand 1S82, 

 and when he has more critically observed the iusectin the planthe will cor- 

 roborate all that I have written in the way of facts and c;bservations and 

 agree with me that they lully justify the opinions expressed, which may 

 be summed up thus: 



I. Yucca by the very nature of its flowers is normally incapable of 

 ■«elf-fructification. This is essentially true of the dehiscent species which 

 'iccur in the Northern States. 



II. Pronub'i ynccasella is modified in the female so as to enable her 

 to gather the pollen and to fructify the fruit. Careful observation of the 

 insect itself proves that function is here, as elsewhere, rightly correlated 

 with structure and that not a solitary ease of Yucca fructification bv an- 

 '^»ther insect has ever yet been witn ss<d by any one. 



III. The only exception to the fruiting of Yucca without the help 

 <^{ Pronuha is in A. uhii/olia, and while I have admitted the possibility of 

 exceptional pollination by other insects, the fact remains that none have 

 been observed to have the power. I have shown that self-fertilization 

 may take place with this species and that its exceptional fruiting where 

 /'roffw^o is out of the question is more likely due to such fertilization 

 than to the agency of other insects; because, from personal observations, 

 both published an.i partly unjiublished, over the whole region in which 

 the plant indigenously grows, I know the fruit to be abundant where 

 Pronuba occurs and that it then always Ijears evirience f»f her presence: 

 whereas where she does not occur the fruit is stant and very excep- 

 tional. 



IV. That a certain variable pf-porf'on of ^■ucta c-af>suies develop 

 without trace </ the Pronuba larvae is explained by the fact that f)ollina- 

 lion, as 1 have shown, often takes place without successful ovipositiou. 

 <»r that successful hatching of the eggN or development of the larvae mav 

 l)e prevented from various causes. I'hat pollination is rare without 

 oviposition is proved In- my direct ol»serv!»tions of the act.s and bv the 

 tact that comparatively few perfect ca[>sules ^or those which do not show 

 ihe evid' n( ir of (ivi[)osi[i( ,ii I arc to Ik: t( .iMi.I c-v<-ii \y].j-r.- .,r) Iar\'e have 



