Inheritance of thu luplandra-iorra of Digitalis purpurea L. 265 



genetic science render it very probable that this remarkable deviation 

 from the ordinary form of Digitalis originated by a sudden mutation 

 involving the loss of one of the genes necessary to the production 

 of the normal type. This conception leads one to contemplate the 

 normal type as the resultant of an aggregate of genes of quite un- 

 surmizable nature and number. How many "abnormal" strains are 

 potentially present in such a species, each derivable from it by the 

 loss of one or more genes, cannot be guessed, but the number may 

 be large, and there may be many other genes quite as capable of 

 independent loss, but whose loss would prove fatal to the derivative 

 genotypes. Mutations resulting in the loss of such essential genes 

 can not usually be demonstrated, of course, and even in those cases 

 in wliich genes are lost which are necessary to sexual reproduction, 

 but which are not essential to successful vegetation, giving rise to 

 sterile forms which may be vegetatively reproduced, it may be im- 

 possible to determine whether the new form is due to a genotypic 

 change or onl}' to a somatic derangement of some sort. Due restraint 

 should be exercised, therefore, in calling such sterile individuals of 

 any species, "mutants", since their genotypic character in relation to 

 the parent form can not be demonstrated. 



As we have seen, the heptandra-loxra. of Digitalis purpurea has 

 been discovered independently a number of times. As it is wholly 

 worthless as a garden variety and would never be grown except as 

 a curiosity or for scientific study, it is fairly certain that at each 

 recorded discovery it has been derived from the normal form. 

 WTiether its first recorded appeal ance in de Chamisso'S garden or 

 any of its later appearances among seedlings of the normal form, was 

 the result of an immediately antecedent mutation, can not be known, 

 for its status as a Mendelian recessive would make it possible to 

 account for all of its occurrences, — widely separated as they have 

 been both in space and time, — as the results of the normal progress of 

 Mendelian inheritance following a single original mutation of unknown 

 antiquity. If the several recorded origins of the heptandra variety 

 have not been due to repeated mutations, this case affords a pretty 

 illustration, albeit a somewhat artificial one, of an important principle 

 which I pointed out several years ago'), namely, that recessiveness, 

 instead of being a handicap as some have supposed, is in reality a 

 great advantage to a new form which is in any way less well 

 adapted to its environment than its parent. This advantage lies in 



') Science X. S. 25; 59c — 591, 12 Ap 1907. 



