Prof. Nichokon ^ J. E. Marr — Phylogeny of the Grapfolifes. 531 



liydrothecEe which are much more complex, and which exhibit very 

 marked modifications of the primitive sicula. 



Taking the above facts into consideration, we find reason to con-« 

 elude that the character of the hydrothecas is the most important 

 point to retain in view in separating different families of the 

 Graptolitoidea. Taking any group of forms with hydrothecae of 

 a similar character, we believe that the next most important point 

 to consider as indicating genetic relationship is the "angle of 

 divergence." On the other hand, the number of stipes in the 



Giioup 2. 



BilYOGRAPTUS retroflexus 



TeTRAGRAPTUS denticulatus 



DiDYMOGRAPTUS fasciculatus 



polypary is a character of minor importance ; since we think it 

 can be shown that, separating Graptolites into groups characterized 

 by their hydrothecae, the different groups will in many cases exhibit 

 a series of parallel modifications as regards the number of stipes in 

 the polypary, the older forms of the group being the more complex, 

 and the later forms undergoing reduction as regards this particular. 

 Our views may be best illustrated by reference to forms belonging 

 to the " genera " Bryograptus, Dichograptus, Tetragraptus, and Didy- 

 mograptus, the appearance of which in time, in the above sequence, 

 is indicated by stratigraphical observations. 



