IOS NEW YORK STATE MUSEUM 



arrangement of the pairs of branches, which in Amphigraptus are so closely 

 crowded around the center as to produce the appearance of a radiate 

 arrangement. Besides, they may all become compound as in A. m u 1 1 i f a s- 

 c iatu s and the whole rhabdosome gain in strength and rigidity. Since in 

 Syndyograptus the secondary branches are little diverging, the rhabdosome 

 settles on its side, while in Amphigraptus it rests on its dorsal or ventral 

 side with the branches spreading out on both sides of the main stipes. 

 The path of the formation of this group is possibly shown by the pecul- 

 iar centribrachiate form of Leptograptus flaccid us described and 

 figured by Elles and Wood [1903, pi. 14, 15]. In this an extra stipe or 

 branch is present in the center of the rhabdosome which is described as 

 originating from the sicula or its immediate neighborhood and as being 

 either simple or compound. As to its origin it is stated \op. cit. p. 108] : 



m In these centribrachiate mutations we have 



/ been unable to determine with certainty how 



I this third stipe arises. Sometimes it looks as if 



gj it were merely the prolongation of the apical 

 part of the sicula, but in other specimens the 

 apex of the sicula is clearly visible. It seems 

 va^ S nA 7 ch ep En ef&^oo U d S . DouWy cen! possible that it is an abnormal growth from the 

 tribrachiate form, n at. size initial bud from which the two normal stipes 



of the polypary are developed, or in other words, it is an extra division of 

 the crossing canal. Thus the branching, instead of being deferred to a late 

 stage in the growth of the polypary, as in the pleurograpti, is concentrated 

 in its early stages. 



One specimen of Leptograptus flaccidus var. macer (here 

 copied in text fig. 47) is a doubly centribrachiate individual. In it two 

 branches spring from the center, in another they appear a little away from 

 the center so that one is in doubt whether one should consider it as a cen- 

 tribrachiate form of Leptograptus flaccidus or an Amphigraptus 

 All these phenomena go to indicate a tendency to extra-division near the 

 center of species of JLeptograptus which in some cases produces paired sec- 

 ondary branches and thus would seem to be competent to explain the pecul- 

 iar phenomenon of paired branches in the genera Syndyograptus and 

 Amphigraptus by similar extra divisions of the connecting canals between 

 the successive thecae of the main stipes. 



