108 



SCIENCE 



[Vol. LV, No. 1414 



the factors or genes responsible for inher- 

 itance, the pure species owes its characteristics 

 to the fact that parents contribute chromo- 

 somes of identical factorial constitution and 

 therefore give to the zygote pairs of homo- 

 logous chromosomes with the exception that 

 genes which differentiate sex can only be 

 present in single sets. Expressed in the 

 terminology of the geneticist the pure species 

 is homozygous for all genes responsible for 

 the species' characters other than those of sex, 

 and for sex characters the germ-plasm is 

 heterozygous in either the male or female indi- 

 vidual at least where animal forms are under 

 consideration. The problems of sex deter- 

 mination from the diploid sporophyte genera- 

 tions of plants are not yet fully solved. Aside 

 from the possibilities of factorial mutations 

 and of mutations due to irregularities of 

 chromosome distribution a pure species must 

 develop gametes identical for all genes 

 other than those of sex, or linked with 

 sex, because the homologous chromosomes 

 during the reductions divisions separate from 

 one another. Some authors would strictly limit 

 the term species and accept that definition of 

 Lotsy (1914), "A species is the total of all 

 individuals of the same hereditarj^ composi- 

 tion, forming but one kind of reproductive 

 cell." I cannot agree with this opinion since 

 the definition calls for what is almost an 

 abstraction in higher animals and plants, the 

 absolutely pure race. 



In contrast to the pure species as defined 

 above is the impure species, the germ-plasm 

 of which in the diploid condition carries dif- 

 ferent sets of genes affecting characters other 

 than those associated with sex. With respect 

 to these genes the germ-plasm is heterozygous 

 and through the reduction division there must 

 take place a segregation of genes with the 

 result that the impure species cannot produce 

 a uniform set of gametes, that is, gametes 

 identical in their germinal constitution. If the 

 diploid germ-plasm is heterozygous for one 

 pair of chromosomes other than the sex 

 chromosomes there would be developed through 

 the separation of the difi'erent chromosomes of 

 such a pair two classes of gametes of each sex 

 provided that reduction proceeds in a normal 



manner. If heterozygous for two pairs of 

 chromosomes there would be developed under 

 normal conditions of meiosis four classes of 

 gametes of each sex, and the theoretical possi- 

 bilities when larger numbers of heterozygous 

 chromosome pairs are present may be calcu- 

 lated by the well known genetical formula (2") 

 when n = the number of heterozygous 

 chromosome pairs. 



The impure species is therefore clearly 

 hj'brid in its genetical constitution but there 

 is this peculiarity in its breeding behavior 

 that it frequently shows little or no evidence 

 of a segregation of contrasting genes. There 

 is in such cases no obvious splitting off of 

 classes through its progeny, but, on the con- 

 trary, the impure species breeds true or nearly 

 true to its type. The true breeding of an 

 impure species must be due to the fact that 

 only favored types of gametes are able to 

 produce in conjugation vigorous zygotes 

 capable of successful development. Further- 

 more, such favored gametes must carry be- 

 tween them those genes which in combination 

 will reproduce the impure heterozygous 

 germinal constitution of the parent stock. 



It is well understood from various plant 

 material that the failure of a hybrid to pro- 

 duce a diverse progeny may be due to irregu- 

 larities at a number of different points in the 

 life history. The death, the sterility, or the 

 failure of maturation of classes of gametes 

 will eliminate the possibilities of development 

 of whole groups of segregates. Even when 

 viable classes of gametes are formed some may 

 leave no progeny because in conjugation they 

 fail to produce zygotes able to develop a suc- 

 ceeding generation. In plants the length of 

 style, or the nature of its tissues, or of its 

 stigma secretions may operate to check or to 

 limit pollen tube growth or the speed of such 

 growth for some classes of pollen grains and 

 at this point in the life history prevent the 

 functioning of pollen tubes carrying particu- 

 lar types of gametes. Pollen and ovule abor- 

 tion in greater or less degrees is a very com- 

 mon phenomenon and is responsible at times 

 for the elimination of entire classes of 

 gametes. High degrees of seed sterility and 

 the weak germination of seeds express the 



