172 THE INVERTEBRATA 



a universal distribution in the Palaeozoic oceans. A series of life 

 zones may be traced in the rocks which were there laid down, each 

 characterized by a definite assemblage of graptolites, and these may be 

 traced throughout the world. By a careful consideration of these 

 graptolite successions the main line of evolution of the group has been 

 worked out. It is now concluded that actual genetic relationship is 

 best traced by the characters of the hydrothecae. The earlier forms 

 have very simple hydrothecae, but the shape becomes gradually more 

 complex. On the other hand the genera were usually founded on the 

 number of branches or stipes in the colony, such as Bryograptus with 

 many stipes in the Cambrian, Tetragraptus with four in the Lower 

 Ordovician, and Didymograptus with two in Lower and Middle 

 Ordovician. These genera succeed each other in geological age, and 

 so we may suppose that they constitute an evolutionary series. In 

 reality they constitute not one but several series. Thus there is the 

 same type of hydrotheca (which we will call A) in Bryograptus callavei^ 

 Tetragraptus hicksi and Didymograptus affinis, while another type (B) 

 is common to B. retroflexus, T. denticulatus and D . fasciculatus . The 

 genera of graptolites as at present constituted are thus open to criti- 

 cism ; it would be more correct to classify all the species into hydro- 

 thecae of type A as one genus, and those into type B as another. In 

 the genus Monograptus, which is the last and most abundant of the 

 graptolites, though the form of the colony is simple, the hydrothecae 

 vary tremendously, and it is obvious that we have here grouped to- 

 gether the descendants of many different genera undergoing com- 

 paratively rapid evolutionary changes. 



Certain forms, whose relationship is not clear, occur very com- 

 monly at certain horizons in the Cambrian and Ordovician and less 

 commonly in later rocks up to the Carboniferous and are grouped 

 together as *' dendroid" graptolites. It is possible that they are 

 closely related to the Calyptoblastea. They differ from the "true" 

 graptolites in showing polymorphism, the thecae being generally 

 interpreted as having enclosed feeding individuals (corresponding to 

 the thecae of the true graptolites), gonozooids (or perhaps protective 

 individuals) and budding individuals. 



Class SCYPHOMEDUSAE (SCYPHOZOA) 



This class contains the common jellyfishes of temp>erate and colder 

 seas, some of which are of extraordinary size, like Cyanea arctica^ 

 the diameter of whose disc is a couple of yards. 



The simplest type of Scyphomedusae is found in the division 

 known as the Stauromedusae, two members of which, Haliclystus and 

 Lucernaria (Fig. 130), are not uncommon on the British coasts. 



