1915] Goodspeed: Parthenogenesis, Parthcnocarpy, PJienospermy 267 



the relatively small amount of viable seed that did mature, to errors 

 in technique. I have attempted to show that the experiments were 

 planned and carried out with a full knowledge of the sources of 

 error and with every effort to guard against their becoming operative. 

 Where contamination was suspected to have occurred in the short 

 interval between the act of castration or mutilation and the bagging 

 of the treated bud, it. as noted above, was discarded in all but a few 

 cases. In these few cases, bags were at once put on and in table 1 is 

 shown the result in the three instances in which the fruits matured — 

 i.e., plant 11, 10/30, plant 14, 9/29 and plant 25, 9/26. In one case 

 nothing but phenospermic seeds were formed, and in the other two hun- 

 dreds of normal viable seeds. This simply means that if the viable seed, 

 which was produced following apparently unimpeachable treatments, 

 was due to chance pollination by wind, or the shaking of neighboring 

 flow^ers' pollen into the bud, larger amounts of viable seed should 

 have been present than were actually found in the nine unsuspected 

 cases of seed production. I cannot feel that the instruments were 

 non-sterile in as many as nine distinct cases. But apart from these 

 considerations, reference to Hartley's paper {loc. cit.) will give infor- 

 mation based upon experiments concerning the effect of premature 

 pollination which must correspond to the general observations of 

 anyone who has been concerned in hybridization experiments with 

 tobacco. Hartley found, that of buds plentifully pollinated with 

 mature, fresh pollen from the same plants, those pollinated more 

 than one day before opening matured 4 per cent of fruits which 

 contained no viable seed, and those pollinated one day or less than 

 one day before opening, matured 86 per cent of fruits which con- 

 tained viable seed. It is thus rather inconceivable that chance pol- 

 lination of small, castrated buds could have resulted in fertilization 

 in such a distinct number of cases. I do not think that light would 

 be thrown upon this particular point by delaying this report until 

 the maturity of the seedlings produced from the parthenogenetic 

 seeds, because the possibility that pollen of a foreign species was 

 present and thus that plants of a hybrid nature and appearance 

 should result, is negligible. It may, in this connection, be stated 

 that cross-pollination of N. Tahacum "JNIaryland" and other N. 

 Tabacum varieties made on "Nic. tahaccum Cuba" have given norm- 

 ally filled capsules of viable seed. 



The facts taken together seem to indicate that three stages are 

 to be observed in the extent to which "Nic. tahaccum Cuba" will 



