116 DR. J. STEPHENSON ON THE MORPHOLOGY, CI ASSIFICA'JION, 



from the ineganephi^idial to the micronephridial ; and that in 

 the prostates, from the tubiil;ir to the racemose form ; in some 

 cases we have changes in the gizzard, in the direction of reduction 

 or reduplication. Especially, in this subfamily, the first three 

 systems are the important ones ; in the majority of genera the 

 gizzai-d remains the same, and the genera are defined by the 

 condition of the setfe, nephridia, and prostates. Since classifi- 

 cation represents relationships, or oiight to, as closely as 

 possible, and since these three systems are those which have been 

 affected in the course of evolution, it is these which form the 

 basis of our generic definitions.* 



Without going into detail, I may refer, in amplification of the 

 foregoing, to two other subfamilies of the same large family. 



The Octochsetinse, like all the subfamilies of the Megascolecidfe, 

 are ultimately derived fi'om the original Notiodrilus form. The 

 first change is the breaking up of the nephridia ; the whole 

 subfamily is therefore micronephridial. In two of the branches 

 of the subfamily the ircrease in the number of the setae takes 

 place. In one genus we have a doubling of the gizzard 

 {Etidichog aster, which I place in the Octochaetinae, cf. No. 1 of 

 the present series). When changes in the male organs occur 

 they take a different line from those which characterize the 

 Megascolecinse ; the '' microscolecine rechiction " which is found 

 in certain genera, consists in the disappearance of the posterior 

 pair of prostates and the translation forwards of the openings of 

 the vasa deferentia to join the anterior prostatic pores on 

 segment xvii. 



The Trigastrinae, like the Octochpetina?, are a much smaller sub- 

 family than the Megascolecinas. Tlie first change in the Notiodrilus 

 a^ncestor along this line is the doubling of the gizzard ; subse- 

 quently we may have the breaking up of the nephridia and the 

 microscolecine reduction of the male apparatus. The only other 

 development is the appearance in certain genera of calciferous 

 glands ; the genera are therefore defined in terms of the gizzards, 

 nephridia, male apparatus, and calciferous glands. 



In these subfamilies we thus see the same thing — evolution 

 proceeds along a, few definite lines, in a definite direction in each 

 system. Sometimes one, sometimes another system is the first 

 to stai-t evolving, sometimes one, sometimes another follows — 

 and so we get a number of combinations, which characterize the 

 different genera. 



* I do not mean to say that the diao;noses of the genera of Megascolecidffi, as we 

 usually have them, are confined to the systems mentioned, thoup;h there is a 

 tendency so to limit them ; compare, for example, the diagnoses of the Megascolecinie 

 given by Michaelsen in the Tierreicli (6) with those by the same author in 1907 (9). 

 A dingno.sis usually contains an element of description, statements which liold good 

 for all hitherto known individuals of a species, or species of a genus, but which are 

 not necessary parts of our conception of the species or genus. Thus we often see 

 recorded the description of species which necessitate the widening of the current 

 diagnosis of a genus. 



