HISTORY OF RESEARCH. cxxi 



another — whether the different individuals have been developed by budding from 

 the coenosarc, or whether they are developed from, and connected with the one 

 immediately preceding— must be considered to begin already in the sicula, even if 

 it has been convenient, for the sake of description, to distinguish one part as the 

 'connecting canal.' " 



The " connecting canal," as defined by Holm, is the canal which " arises almost 

 simultaneously with the left theca and the common canal for the left half of the 

 polypary," and which " crosses the dorsal side of the sicula and gives rise to 

 the third theca and the common canal for the right half of the polypary." (This 

 canal is not the same as that for which Tornquist used the term " connecting canal," 

 and Tornquist has later proposed the name " crossing canal " for Holm's structure.) 



The Virgula. — His observations lead him to the conclusion that " a virgula 

 corresponding to that in Diplograptus and Monogra/ptus cannot occur in the 

 Dichograptidas," at any rate "embedded in the dorsal side of the branches." 

 Even in those cases in which the " sicula is embedded in the polypary, a 

 virgula need not of necessity be present," and Holm fails to find any trace of one 

 in Phyllograptus. 



He points out that the " cylindrical chitinous thread which originates as a 

 result of growth within the apertural end of the sicula " in Diplograptus, etc., as 

 described by Wiman, " stands evidently in no relation whateA^er to the real virgula, 

 but may be regarded as an apertural spine." And he draws especial attention 

 to the fact that "the presence of a virgula has hitherto been considered as the 

 main character of Graptolites " (Rhabdophora), "although such was never 

 described or expressly mentioned except in the groups Diplograptidas, Mono- 

 graptidse, and Retiolitidas." 



The structure of Didymograptus minutus, D. gracilis, and D. gibberulus is 

 described in detail. 



The genus Tetragraptus, of the development of which little was known, is 

 shown by Holm to pass through the same early stage as a Didymograptus 

 (" Didymograptus stage"). The four stipes arise by a "direct splitting of the 

 common canal by a vertical wall " on each side of the connecting canal. This 

 structure is worked out in specimens of T. Bigsbyi. 



The development of the genus Phyllograptus is proved to be practically identical 

 with that of Tetragraptus, but the branches, instead of having " four independent 

 periderm walls, form a single, cruciform, four-winged, longitudinal septum." The 

 sicula is embedded in the polypary, but no virgula has been detected. 



This memoir is illustrated by excellent figures. 



The second important paper, published in 1895, was by Wiman, and was 

 perhaps even more far-reaching in its results. 



The author commences with an account of the methods adopted by him for 

 preparing the specimens examined. 



