MUTATION. 171 



or worse, vegetatively propagated material, or Oenotheras, 

 the geno-typic purity of which is not investigated. An extreme 

 case of the misapplication of this term is certainly Blaring- 

 hem's use of the word, speaking of sheep and guinea-pigs! 



There can be only one exception to the rule that test-matings 

 are necessary to show the homozygosity of the material, to 

 conclude that the sudden production of a novelty is a real mu- 

 tation, a real spontaneous change of genotype, acquisition or 

 loss of a gene. And this exception is the case in which we know 

 that a certain individual must, because of the very mode of its 

 origin, be pure for any gene it contains. If an animal or plant 

 develops from one single gamete we think it could not possibly 

 be otherwise than homozygous. 



Really parthenogenetic individuals must be pure, and they 

 should furnish the only irreproachable material for experi- 

 ments with pure lines. Such animals as in the experiments of 

 Loeb developed out of real fertilizable eggs of sea-urchins and 

 frogs, cannot be otherwise than homozygous. 



In 1912 we started a set of experiments with Squashes and 

 Marrows. In these plants it is possible to obtain fruit contain- 

 ing good seed from unfertilized female flowers. Our method 

 was to close the female buds with lead-wire, removing the male 

 buds on all the plants every evening. We performed several 

 hundred cross-fertilizations between widely different forms. 

 The hybrids, grown in 1913 proved to be in every instance 

 different from the mother- form, thus proving the absence of 

 apogamy. Reciprocal hybrids were in every instance identical. 

 On these hybrid plants we closed the female buds. In doing so, 

 we found that several hybrids would never produce any fruit 

 from closed buds, though they would later in the season pro- 

 duce numerous fruit from fertilized flowers. Other plants pro- 

 duced parthenocarpic fruits, apparently normal, but void of 

 seeds. Still others yielded good fruit with empty seed, and fi- 

 nally, there were some hybrids which produced normal fruit, 

 full of viable seeds. It is to be noted, that in no instance did we 

 find a fruit with only a few good seeds, such as would probably 



