Vererbung, VariatioiJ, Mutation. 457 



seeds in Spermatophytes or vessels in the wood of Angiosperms, must be be- 

 cause of their supreme importance for survival. The study of phylogeny explains 

 certain facts which are un explainable by the selection hypothesis, e. g. the number 

 5 in echiuoderm structure or the number 3 in hexapod insects. It is very doubt- 

 ful if the constancy of the vertebral cohinin in vertebrates or of feathers in 

 birds in due to the supreme importance of these characters in the struggle for 

 existence. The "conservative" character is only useful when associated with 

 many other "variable" characters. Thus a fern and a seed plant differ most fun- 

 damentally in their method of reproduction, ,but in competition between them it 

 is not this difference primarily which decides the result, The theory of natural 

 selection can not, therefore, be regarded as a satisfactory explanation for struc- 

 tural conservatism. Darwin called attention to the fact that "the physiological 

 importance of an organ does not determine its classificatory value". His defence 

 was that "the importance, for Classification, of trifling characters mainly depends 

 on their being correlated with other characters of more or less importance". 

 But it has been found impossible to discover such characters and use them for 

 Classification. 



There are certain definite categories of conservative characters, those of 

 number, relative position and general plan. Plastic or inconstant characters are 

 chiefly those of size, colour and texture. Again, the more primitive families of 

 plants and animals are more variable in their diagnostic features than are the 

 more advanced. Thus the Compositae and Orchidaceae have a stereotyped floral 

 plan, while in the more primitive Rosaceae and Cyperaceae the floral type is 

 more various. Early stages in ontogeny are also believed to be less variable 

 than later ones. Selection, therefore, instead of making for fixity tends to place 

 a premium on versatility and variability of type. 



Organic evolution is dependent upon two opposing factors (l) that of pro- 

 gressive fixation, which tends to increase during evolutionary advance, and (2) 

 that of natural selection, which tends to increase the variability of characters 

 which are important for survival, by eliminating individuals whose characters 

 have become too fixed for the necessary degree of adaptability. Natural selection 

 is not concerned with harmless and trivial characters, which consequently tend 

 to become conservative. Gates. 



1213) Alierbacli, F., Die Variationskurve in der Biologie. In: Zeitschr. 

 ind. Abst.-Vererbsh, Bd. 11, Heft 1/2, S. 18—38, 1913. 



Ausgehend von allgemeinen Studien über graphische Methoden und Gesetze 

 werden vom Verf. die in der Biologie gebräuchlichen statistischen Methoden 

 einer Kritik unterzogen unter Hinweis darauf, daß die zur Gegenüberstellung der 

 empirisch gefundenen Variationskurve von den Biologen als theoretische Wahr- 

 scheinlichkeitskurve benutzte sogenannte Binomialkurve nur in einem ganz 

 bestimmten Fall wirklich die wahrscheinliche Verteilung einer großen Anzahl von 

 Werten auf die Individuen ausdrückt, dann nämlich und nur dann, „wie die mög- 

 lichen Werte der betreffenden, statistisch zu erfassenden Größe einen Bereich 

 umspannen, der nach beiden Seiten hin im Prinzip unendlich ist, d. h. von minus 

 unendlich bis plus unendlich sich erstreckt; praktisch genommen aber mindestens 

 nach beiden Seiten sehr groß, und zwar wesentlich gleich groß ist". 



Dieser Fall tritt in der Biometrie im Prinzip überhaupt nicht auf und prak- 

 tisch nicht mit genügender Annäherung, Es handelt sich in der Biometrie viel- 

 mehr fast stets um Größen positiver Natur, die einseitig begrenzt, oder auf einer 

 Seite enger begrenzt sind (Körperlänge des Menschen, Zahl der Blumenblätter). 



