Vol. 54.] TAUNA OF THE SKIDDAW SLATES. 629 



(Hall) associated with Bryograptus ramosus var. cumhrensis, show- 

 ing that T. Bigshyi must occur in the lower parts of the beds. 



(2) Tetragraptus serra (Brongn.) and other forms are also found 

 associated with the fauna which in Sweden characterizes the zone 

 of Phyllograptus cf. typus and occur on the same slab as the 

 earliest ' tuning-fork ' Didymograpti. 



(3) Dichograp>ti do not occur with this higher fauna at all, and 

 therefore are probably earlier forms. 



The Upper Skiddaw Slates are characterized by the abundance of 

 Diplograptidae and ' tuning-fork ' Didymograpti. 



The results of my work as regards the stratigraphical relationships 

 of the beds are summed up in Table III (p. 530). 



[b] Phylogenetic. 



It has been established as a general rule that graptolites with 

 numerous branches have been gradually succeeded in time by 

 types in which the branching was simpler. This is no doubt true 

 in. a general sense, but I fully agree with Nicholson & Marr 

 that the phylogenetic relationships of the graptolites are not so 

 simple as they appear. These authors, in their paper on the 

 ' Phylogeny of the Graptolites ' (Geol. Mag. 1885, pp. 529-539), have 

 called especial attention to the remarkable resemblances which exist 

 between the species belongiiig to different genera of the Dicho- 

 graptidse, and have suggested that these resemblances may be 

 indicative of genetic relationship. They point out how very difficult 

 it is to understand how the extraordinary resemblances between the 

 various species of Tetragraptus and Bidymograptus, etc., have arisen, 

 if, as is usually supposed, all the species of these genera have 

 descended from a common ancestral form for each genus, in the one 

 case four-branched, and in the other case two-branched. On the 

 other hand, they hold it to be comparatively easy to explain the more 

 or less simultaneous existence of forms possessing the same number 

 of stipes, but otherwise only distantly related, if we imagine them 

 to be the result of the variation of different ancestral forms along 

 similar lines. They urge that, when the graptolites (Dichograptidae) 

 are separated into groups characterized by their hydrothecae, 



(«) The different groups exhibit a series of parallel modifications 

 as regards the number of stipes in the rhabdosoma ; 



(h) The older forms of the group are more complex, and the 

 later forms undergo reduction in the number of stipes. 



They distinguish nine such groups in the family Dichograptidae, 

 and select as their first and typical group the series represented by 

 Bryograptus Callavei, Tetragraptus Hicksii, and Bidymograptus 

 affiyiis. 



In the course of my own work among the Skiddaw Slate grapto- 

 lites, I have also been greatly impressed by these remarkable 

 resemblances between species of different genera, and, like these 

 authors, I feel confident that such forms will be found eventually 



2o2 



