86 JOHAN KIÆR. M.-N. Kl. 



mens. They are more differentiated in Callavia in which they have extended 

 far outwards. In Wanneria and Kjeridfia on the other hand we only find 

 a more or less slight intergenal angle that is often slightly prominent. As 

 far as this structural feature is concerned we should have to make the line : 

 Hohnia, Cal/avia, IVaiineria, KJenilfia. 



With respect to the number of segments, we must assume that 

 throughout the evolution of Mesonacidae we have a gradual reduction, e. g. a 

 regressive development. This is clearly seen in the Oloiclhis line, and must 

 also have effected the development of the other genera dealt with here. 



It is now found that they all have the thoracic segments divided into 

 two sections, first a long, uniform section consisting of large and well 

 developed segments, and than a hinder section with one or two rudimen- 

 tary segments. In the case of Kjendfia it is presumably 17-I-1 (18); in 

 the type of Callavia { Brøggeri Walcott) it is 16 -|- 2 (18), but it increases 

 in the other and somewhat doubtful species to 23, and may decrease to 

 15 1. In Wanneria there are 15 + 2 (17), and in Holiuia 14 + 2 (16). We 

 thus obtain the probable line: Callavia, Kjeriilfia, Wanneria, Hohnia. 



If we regard the form of the pleurae, we find undoubtedly the most 

 primitive stage in Hohnia, the pleurae as in Mesonacis ending in oblique 

 and receding spines. In the other three genera, like the Olenellus line, 

 they are transformed, and more extended into curved, falcate spines. They 

 are presumably most markedly specialized in ICanneria, where the hind- 

 most segment in particular is remarkabl}' extended as in Olenellus. We 

 must probably assume in this case also that a similar biological cause has 

 affected the transformation. 



If in accordance with the above we arrange a series of progressive 

 stages we obtain: Holmia, Kjeriilfia, Callavia, Wanneria. 



Comparing the results that we have obtained by investigating the evolu- 

 tion of these four structural features, it will easily be seen that the lines cross 

 each other, and that we have a marked specialization crossi ng, in the 

 sense deweloped by Dollo and Abel. 



None of these genera therefore appears to have such a phylogenetical 

 relation to each other that it is possible to arrange them in a line or series of 

 stages of development. 



I therefore consider them as groups that have developed in special lines 

 of evolution, and the earlier stages of which are still unknown to us. 



As far as concerns Callavia, the evolution would appear to be made 

 somewhat more clear by the forms recently described by Walcott. His 



1 Walcott, Olenellus etc., page 275 and New Cambrian Subfauna, 1913, page 315. 



