HYDROZOA. GRAPTOLITOIDEA 63 



(1) the first hydrotheca (fig. 17, l) and (2) a tubular body 

 known as the crossing-canal (c); the latter grows across 

 the sicula and gives rise to the second hydrotheca (2) 

 which is on the side opposite to the first hydrotheca. 

 From each of these two hydrotheca? a stipe or branch is 

 developed, owing to the fact that each hydrotheca gives 

 rise by budding to another hydrotheca and in this way 

 two continuous linear series are formed. More complex 

 branching (as in Tetragraptus, Dichograptus) is produced 

 when one hydrotheca buds off two hydrotheca? instead of 

 one. In Diplograptus (fig. 18 C) the development is 

 similar to that of Didymograptus but the crossing-canal 

 is reduced in size and the hydrothecse grow up the virgula 

 instead of in the direction of the apertural end of the 

 sicula ; also the hydrothecse are budded off on either side 

 alternately, so that the second hydrotheca (2) is on the 

 side opposite to the first (1), and the third (3), which is 

 budded from the second, is on the same side as the first 

 and overlies it. This alternate arrangement may continue 

 throughout the development of the polypary, but frequently 

 a septum appears between the two rows of hydrotheca?, 

 and thenceforward each hydrotheca arises from the pre- 

 ceding one on the same side. In Monograptus the 

 hydrotheca? (fig. 18 B) arise on one side only of the sicula. 

 Owing to the fact that the soft parts of the graptolites 

 are entirely unknown it is difficult to speak of their 

 affinities with any degree of certainty. It seems probable, 

 however, that they belong to the Hydrozoa ; Allman and 

 others consider them to be closely related to the Calypto- 

 blastea, especially to such forms as Sertularia and Plumu- 

 laria, with which they agree in the general characters of 

 the hydrotheca? and common canal, and perhaps also in 

 the possession of gonangia. But they differ in some 



