266 HORTICULTURAL SOCIETY OF NEW YORK. 



remain on the spike until they bloom, and after some experience one will 

 get on quite well without them. The treated flowers were generally de- 

 prived of their anthers just as the spike made its appearance above the last 

 leaf. 



The second part of the operation is to fertilize the pistils of the treated 

 flowers when ready to receive the pollen. The length of time between these 

 two operations will vary according to maturity of flowers when the anthers 

 were removed and also with conditions of weather. Successful crosses were 

 made when both operations were performed on the same day and also when 

 four days elapsed between the two operations. 



Much of the success or failure of the experimenter depends upon the cer- 

 tainty with which he is able to furnish the ripening pistil with good healthy 

 pollen. On this point the writer made an observation while at work in the 

 field during the season of 1884, which ever afterward solved the question of 

 securing proper pollen. This observation was the actual blossoming of one 

 of the flowers on a head under treatment. The glumes were seen to part with 

 a slow but plainly visible motion, the palae also parting with them and ex- 

 posing the enclosed anthers and pistil to view. The width of the opening was 

 not measured but was judged to be between one- and two-sixteenths of an 

 inch. The filaments which support the anthers rapidly elongated, thrusting the 

 latter through the opening when they immediately turned downward and shed 

 their pollen. 



(Noting that the pollen was shed suddenly just as the anthers were thrust 

 out of the parts enclosing the flower, it occurred to me that that was the time 

 to take the pollen for the work. This was afterwards followed except that I 

 learned how to hasten the process of ripening by plucking heads on which 

 flowers were nearly mature, drawing back the outer glumes and palets on one 

 side and exposing the anthers to hot sunshine. The pollen was gathered in a 

 small dish and applied with a camel's hair pencil to pistils of the flowers pre- 

 viously prepared, opening the glumes as already described, by means of the 

 small forceps. The glumes and pales of worked flowers open when the pistil 

 is ripe just as though the anthers had not been removed, hence pollen may 

 reach a castrated flower by natural agencies.) 



I do not wish to pass the operation of blooming as observed in numerous 

 instances during my later work without noticing it more particularly. The time 

 occupied by the flower in opening is from one to two minutes, the anthers can 

 be seen to immediately rise out from the opened flower and shed their pollen, 

 and strictly speaking I consider the period of bloom lasts about five minutes, 

 certainly does not exceed ten. As soon as I saw the copious amount of pollen 

 poured out among the spikelets on the head the fact was at once impressed 

 upon me that the wheat flower was not of necessity strictly self-fertilizing. 

 Any other flower on the same or adjacent head which might be open at this 

 time is very liable to receive pollen and thus at least its own pollen be assisted 

 in the act of fertilization. With the question of natural cross-fertilization in 

 view I was led to examine more carefully the flowers and found that in a 

 great many instances the filaments of the two-branched plumose stigma extend 

 out laterally, far enough to reach quite through the enclosing parts, and in my 

 estimation making it possible for them to receive pollen from the outside if by 



