British Tuhificidas. By Rev. H. Friend. 269 



emphasized the value of the nervous system and the ganglia of the 

 head ; another has dwelt on the vascular arrangement, the shape 

 of the segmental organs (nephridia), or the presence or absence of 

 penial setse and a penis-sheath. The building of a tube was at first 

 regarded as characteristic, but it is now known that practically all 

 the true Tuhijicids plunge their heads in the mire and sway their 

 tails in the water, thus making a kind of tube ; whereas real tubes 

 are usually fabricated when the worm is injured, in danger, or pass- 

 ing through some special phase of its life-history. 



I propose to take, first, such genera as are easily differentiated, 

 and leave over the more complex groups till we have become familiar 

 with the main characteristics. It will be easy afterwards, when our 

 knowledge is a little more complete, to throw the whole into a 

 suitable order and classify the genera on a definite plan. And as 

 experience teaches me that the set* are the most constant, if not 

 by any means the most important characteristic, I shall lay con- 

 siderable stress on the number, shape, variety, and disposition of 

 these external features, and show how by their more careful study 

 it will be possilde for us in the future to avoid many of the errors 

 into whicli earlier observers have been led. 



GROUP I. — Capilliforx Set.e (Haarborsten) Absent. 



A. — Forked Setse only present. 



Beginning with those whose set?e are all of one kind, and will 

 therefore afford a ready means of identification, I select the genus 

 Zimnodrilus for our first consideration. Beddard's brief diagnoses 

 of the species are unfortunately valueless, because they omit many 

 of the most essential characters. Michaelsen has extended the 

 descriptions somewhat, but even he does not always make the 

 details clear. Hence the need of a fuller elaboration. 



I. Genus Limnodrilus Clap. 



This is, perhaps, the best defined genus in the family. Forked 

 setae only are present ; these in four bundles which vary in number 

 and shape in the different species. Contractile hearts in segments 8 

 and 9 as a rule. A chitinous penis-sheath usually * present, whose 

 length always exceeds its breadth. Male efferent apparatus carry- 

 ing a prostate. Spermathec^ in the 10th segment, usually with 

 spermatophores. Male apparatus opening in segment 11. 



Fresh-water Annelids, found in Europe, North America and 

 Japan. Beddard (1895) allowed 10 species : Michaelsen (1900) 

 gives 9 with 3 others which are uncertain. Of these 2 are entered 



* See, however, Pointner's article, and the^note on Limnodrilus papillosus 

 Friend, infra p. 276. 



