a Family of Graptolites. 21 



carried too far. This opinion also seems to be held by Tullberg 

 (Skane's Graptoliter ii.). It would be better to regard them as sub- 

 genera of the two great genera Didymograptus and Tetragraptus. 



All the many-branched Dichograptida? may be conceived as origi- 

 nating from a Tetragraptus by repeated furcation. A division, 

 repeated in all branches, produces eight-branched forms ; further 

 complete furcations give 16-branched forms and so on. Hitherto 

 the 8-branched forms constituted the genus JDichograptus, Salt., and 

 those with 16-32 branches the genus Loganograptus, Hall. But in 

 considering the various forms with 8 and more branches, we are at 

 once struck with an important contrast to those with 2 and 4 branches. 

 The latter, when perfectly preserved, invariably possess only 2 and 

 4 branches ; but in the many-branched forms a remarkable variation 

 occurs in the number of branches ; more and more frequently the 

 full number is either not developed or exceeded, imperfectly or 

 abnormally developed individuals become more and more numerous, 

 so that the number of branches becomes of less and less value as a 

 distinctive character. Thus in the multiramose species of the genus 

 Clonograptus, Hall, the number of branches oscillates between 40 

 and 60 for the same species. The number of branches becomes finally 

 so complicated as to be quite unimportant (e.g. Clematograptus mul- 

 tifasciatus, Hall). Thus the genus Loganograptus, Hall, with 16-32 

 branches appears to be quite arbitrary and artificial, and must be 

 suppressed, unless we are prepared to establish a genus for every 

 many -branched species. It has been proposed to refer all species 

 with more than four branches to Dichograptus, Salt., which has 

 hitherto included only 8-branched forms. But upon this the author 

 would put a limitation, as there is in the many-branched forms a 

 striking difference. " Thus either all the points of furcation are 

 situated in the immediate vicinity of the sicula (e.g. Dichograptus 

 Kjerulfi, Herrm., Fig. 8), while the more distant parts of the branches 

 remain undivided ; or there are greater distances between the indi- 

 vidual points of furcation, so that the branches may divide through- 

 out their whole length, and the points of furcation may depart far 

 from the centre of the hydrosoma (e.g. Fig. 9)." For the latter 

 forms Hall has proposed the name of Clonograptus, which may 

 stand, and thus we get two homologous series forming the genera 

 Dichograptus, Salt., and Clonograptus, Hall, each including forms 

 with 8, 12, 16, and more branches. 



Genus XIV. Dichograptus, Salt, (modified). 

 Hydrosoma bilaterally symmetrical. From the broad end of the 

 sicula, which is directed upwards, issue two short oppositely diverging 

 branches, which fork. Of the four branches thus produced, all or a 

 part may again divide dichotomously once or several times at very 

 short distances, so that more than four equivalent, simple branches 

 are produced. The proximal parts of the hydrosoma are frequently 

 enclosed in a central disc. Hydrothecae as in Didymograptus, M'Coy. 



