56 F. H. Mc Learn — Lower Ordovician 



in smaller sizes is repeated in the ontogenetic development of 

 all the species. The nema cannot be demonstrated in the 

 specimens now available, although one or two individuals of 

 P rotistograptus corrugatus show what may be the beginning 

 of such a process (see tig. 2e). The size varies, but a maxi- 

 mum length of 10' ,Hn and a maximum breadth of l-2 mm may be 

 attained. 



Remarks— The genus is probably to be regarded as the 

 prototype of the Graptolitoidea, just as Paterina occupies a 

 similar place in the Brachiopoda. The analogy is not an exact 

 one, however, because the sicular cone includes more than the 

 embryonic shell, i. e., more than the equivalent of the prodisso 

 conch of the Pelecypoda, the protaspis of the Trilobita, or the 

 protegulum of the Brachiopoda. On the basis of the morpho- 

 logical studies of Wiman, Holm (1895, p. 3) considers the 

 initial part of the sicula with the thin periderm to be the 

 embryonic shell, and the apertural part with growth lines, 

 apertural spine, etc., to be a later growth. Kuedemann, by 

 comparison with the development of a tubularian hydroid, 

 demonstrates that the initial part of the sicula, the nema, and 

 the disc, all together make up the embryonic shell. In Pro- 

 tistograptus, the presence of a post-embryonic structure, the 

 apertural spine, shows that the genus is the homologue of a 

 sicula, i. e., of the embryonic shell plus some later growth. It 

 is still, however, in the single zooid stage. 



By budding and dichotomy, the rhabdosome would pass 

 through Didymograptus, Tetragraptus, etc., stages. This 

 statement does not necessarily mean that the genera referred 

 to arose in this order, but that the graptolites, at some time in 

 their history, probably passed through similar stages. 



The detailed characters in the various siculse have not, as 

 yet, been demonstrated to be of value in phylogenesis. Tims 

 each phylum may be characterized by a particular kind of sicula, 

 or the sicula itself may take on new characters pari passu 

 with the evolution of the rhabdosome. The sicula is seldom 

 described in detail, so that it is impossible to determine 

 whether its variations in form and structure are of genetic 

 value. It is equally impossible to determine what are the 

 primitive characters of a sicula, but the literature and figures 

 show that they vary in slenderness, shape, and number and 

 shape of the spines. 



It may be inferred that the most primitive sicula would be 

 straight, conical, and without apertural processes, i. e., a mere 

 continuation of the embryonic shell. The sicula of Corynoides 

 fulfills these requirements most closely. Here the sicula is a 

 short or long cone without an apertural spine. This genus 

 possesses other characters that may be interpreted as primitive — 



