Varietäten, etc. — Palaeontologie. 85 



&• ' 



recessives in the ordinary sense and that the commercial growers 

 are right in asserting that a strain may breed true for a while and 

 tben throw rogues. 



The variety E.G. is peculiar in that there occur intermediates 

 of several kinds. These may be graded somewhat arbitrarely thus. 



1) Types-pods straight &c. 



2) Types: occasional pods curved. 



3) Intermediates: leaves and stipules nearly as in type: pods 

 definitely curved. 



4) Rogues: foliar parts somewhat larger than in ordinary rogues: 

 pods curved. 



5) Rogues: foliar parts small and pointed: pods curved. 



The genetical behaviour of types and intermediates may be 

 summarised thus. 



T3^pical plants (class 1 above) often breed true but they may 

 throw any of the aberrant forms though never more than a small 

 percentage of them. 



Plants of class 2 commonly behave like the types themselves. 



Most of class 3, the real intermediates, give a large majority of 

 rogues and so must be quite different geneticalty Irom types of 

 classes 1 and 2. 



The numerical proportions in which the aberrant forms are 

 produced from the types are such as to render it extremely unli- 

 kely that the}'' are expressions of any ordinary factorial System. 

 The interrelations of classes 4 and 5 are equally obscure. 



Returning to the consideration of the varieties as a whole, 

 remarkable evidence was provided by crossings between types and 

 rogues. 



52 families were raised by crossing typical plants with rogues 

 of classes 4 and 5. With the exception of two families, all these 

 families when fully grown were rogues in all respects. The result 

 was the same whether the type was used as maternal or paternal 

 parent. In the juvenile condition however they differed little if at 

 all from the types. One of the two exceptions mentioned above 

 gave rise to plants probably of class 2; the other threw some inter- 

 mediates (class 3). The course of the phenomena is quite unlike 

 ordinary Mendelian inheritance. 



The authors consider that the evidence points to the „type 

 elements", when introduced from one side of the parentage only, 

 being in some way cut out of the germ lineage in the early stages 

 of somatic development and lost. It should be mentioned that the 

 genetic behaviour of the ordinary factors introduced in crosses 

 with rogues is normal. Thus it is only the factors special to the 

 type which are excluded or eliminated. 



In a note added in June 1915 the authors mention another point 

 of considerable importance that has come to light in some families 

 subsequent to those just discussed. A cross bred plant of 1914, 

 booked as a rogue, had a branch somewhat type-like in character. 

 Seed saved from the rogue-like parts of the plant has given all 

 rogues, but the seed of the type-like branch has given some type- 

 like plants in addition to many rogues. W. Neilson Jones. 



Halle, T. G., A fossil Sporogonium fromtheLowerDevonian 

 of Röragen in Norway. (Botan. Notiser. p. 78—81. Fig. 1916.) 



A diagnose is given of Sporogonites exuberans n. g. n. sp. The 



