INTRODUCTION. 



The Calycophorida and Physophorida I should also rank 

 as suborders,, and unite in the single order Siphonophora. 

 The Medusida of Huxley, under which he has ranged the 

 naked-eyed Medusae that have not yet been traced to a 

 Hydroid stock, and those which are known to be developed 

 directly from the ovum, according to the views already 

 stated, should cease to constitute a distinct group. For the 

 Steganophthalmata (Forbes) with Lucernaria, which form 

 the third of the orders of Hydrozoa, Discophora seems to 

 me a better designation than Lucernariada, which has been 

 adopted by Huxley, and which is derived from a strikingly 

 aberrant form. 



In classifying the Hydroida and constructing the generic 

 groups, respect must be had, as emphatically pointed out 

 by Allman*, to both the nutritive and reproductive ele- 

 ments. It is much easier, however, to recognize the 

 correctness of this principle in the abstract than to apply 

 it practically to the work of the systematist; for the 

 affinities suggested by one of these elements are, in many 

 cases, by no means affirmed by the other. 



The trophosomes of two species may agree very closely 

 in character, while the gonozooids are widely dissimilar, 

 and vice versa. To take a striking illustration : the repro- 

 ductive zooid of Corynopsis, a genus which ranks in the 

 family of the Podocorynidce, is identical when first liberated 

 with that of Bougainvillia, a member of the family Atrac- 

 tylida-\. So Syncoryne eximia and Stauridium productum, 

 which are referred to different genera from the dissimi- 



* In bis valuable paper " On the construction and limitation of genera 

 among tbe Hydroida," Ann. N. H. for May 1864. 



t By an error, Bougainvillia is referred, on page 35, to the family of the 

 Eudcndriidce. 



