Historical 13 



the orders. The orders were thus defined : (1) Nereideae feet with 

 retractile subulate chaetae but without crotchets, a distinct head with 

 eyes and feelers, and a protrusible proboscis almost always armed 

 with jaws; (2) Serpuleae feet with retractile subulate chaetae and 

 crotchets, head without eyes or feelers, proboscis not armed with 

 jaws ; (3) Lumbricinae without projecting feet and with chaetae 

 rarely retractile, head without eyes or feelers, and without jaws ; 

 (4) Hirudineae without locomotor chaetae, characterised by the 

 presence of a sucker at each extremity and by having eyes. The 

 families were arranged in the orders thus : 



I. With chaetae for locomotion. 



1. Order Nereideae. Families Aphroditae, Nereides, Eunicae, Am- 



phinomae. 



2. Order Serpuleae. 



(1) Gills none or few ; if present situated on the anterior segments 



of the body ; feet of several kinds. 



Families Amphitritae, Maldaniae. 



(2) Gills numerous, not on the anterior segments of the body ; feet 



of one kind. 



Family Telethusae. 



3. Order Lumbricinae. Families Echiuri, Lumbrici. 



II. Without chaetae. 



4. Order Hirvidineae. Family Hirudines. 



This is the first list of Annelids published from which all 

 Mollusca are excluded. 1 It is also of interest from our special point 

 of view, as the family Telethusae was formed for the reception of the 

 single genus Arenicola. 



Savigny's classification is more natural than any of its pre- 

 decessors, as it is based not on a single feature, but on a group 

 of characters. More than forty years afterwards Prof. Ehlers con- 

 sidered it to be so excellent that he adopted it, with only a slight 

 change, in his monograph on " Die Borstenwurmer," and the 

 classification in use at the present day is based on an extension of 

 the principles of Savigny's system. In addition to his fundamental 

 improvement of the classification of Annelids, Savigny also carefully 

 delimited the known genera and -founded more than a score of new 

 ones, most of which remain valid. 



Latreille 2 adopted Cuvier's classification in a modified form, 



1 Savigny observed that Dentalium was not an Annelid, and he also 

 excluded L'Arrosoir and Siliquaria. 



2 Families nat. Begne Anim., Paris (1825). 



