GENETICS: T. H. GOODSPEED 
343 
ing off the stigma in the bud without the removal of the anthers. The 
crushed tip of the style in such cases dries down considerably before 
the flower matures or pollen is shed. The last two types of treatment, 
(2) and (3), will be referred to as 'mutilations.' In addition a few 
decapitation experiments were attempted (cf. Thomas, I.e.). The num- 
ber of simple emasculations far exceeded the other two types of treatment. 
The length of every bud treated was noted by a measurement from 
the point of union of calyx and pedicel to the tips of the folded corolla 
lobes. The maximum length of unopened buds was 49 mm. and, with 
the exception of a few late flowers, the anthers do not open until the 
corolla lobes begin to open back. The number of the flower parts was, 
also, determined for each bud on the supposition that four-parted and 
five -parted flowers might give different results following emasculation 
or mutilation. Finally, the records were so kept that the position on 
the plant of a given treated bud can be determined from the field notes — 
i.e., whether it occurred on the terminal inflorescence, on one of the 
upper 'bald suckers', or on one of the lower leafy laterals. More than 
one bud was often treated and bagged on a single inflorescence but in 
most cases such a group of buds were all of approximately the same 
length. 
To obviate the possibility that grosser errors of technique could 
affect the results of the various treatments, the corollas of treated 
flowers were carried along with the seed bags and are, in most cases, 
still available for examination. If, in castration, an anther is over- 
looked the fact is at once noted when the withered corollas are exam- 
ined both at the time of cleaning and at the time of sowing the seed 
produced. In only one case among the 800 treatments were anthers 
overlooked and a considerable amount of viable seed resulted. This 
check upon the results of castration and mutilation experiments is, 
to my mind, imperative. Willingness to discard all treatments concern- 
ing which there is the slightest doubt as to purity, is the only security 
obtainable in the finer details of performing the emasculations or muti- 
lations. Despite the sterilization of instruments a bud was discarded 
if the forceps touched the stigma during any operation. Similarly, 
though only the middle one-third of the corolla tube v/as opened to admit 
the forceps, the too-close proximity of open flowers on neighboring in- 
florescences at the moment of treatment caused the discarding of the bud 
being operated upon. In addition to the 800 castrations and mutila- 
tions considered unimpeachable, 150 doubtful treatments were thrown 
away. The results, in the case of a few doubtful treatments which 
were saved, showed the necessity of such rigid discrimination. 
