LEWIS, — POLYCH^TE ANNELIDS. 229 



body wall resulting from the contraction of the animal. A narcotizing 

 mixture of about 5 volumes of 95% alcohol to 95 volumes of sea water 

 gave the best results. A fuller account of the methods employed for 

 studying different structures will be given under the separate topics. 



3. Leydig's Fibres. 



I have selected the name Leydig's fibre for a long known structure in 

 annelids described under several different names ; the principal ones, 

 besides that which I have chosen, are neurochord, giant fibre, central 

 canal, and neural canal. The name Leydig's fibre is adopted because 

 Leydig was the first writer to give an accurate description of this 

 structure ; he also ascribes to it the nature which the most thorough 

 and careful of recent investigations shows to be the most probable. 



a. Historical Review. 



These fibres have been studied by many writers, and the literature 

 upon them is extensive. Bibliographical lists of considerable length, 

 arranged chronologically, have been given by Spengel ('81, p. 41), Eisig 

 ('87, p. 476), and Friedlaender ('89, p. 206 ; '94, p. 662). 



But in spite of the fact that so much attention has been given to these 

 structures, the most widely divergent conclusions have been drawn re- 

 garding their morphology and function, the latest papers upon the subject 

 being no more harmonious than were the earlier ones. At least five 

 different theories regarding the Leydig's fibres in annelids have been 

 strongly urged. 



First, that they are true nerve fibres of large size. This view was 

 first set forth by Leydig in 1864, and it has been consistently supported 

 by him in several subsequent papers. He declares these structures to be 

 large medullated nerve fibres, — '* riesige, dunkelrandige Nervenfasern," — 

 and regards them as similar to the medullated nerve fibres of verte- 

 brates. This theory received confirmation from Spengel ('81), who was 

 the first to show that these fibres were the processes of ganglionic cells. 

 Since, in the case of the annelid which he studied the number of longi- 

 tudinal fibres remains tolerably constant and is much less than the 

 number of giant cells, he believes that the larger fibres — ''grosseren 

 Rohren" — have arisen through the union of the direct processes of 

 these ganglion cells. 



Secondly, the view defended by Claparede (first in 1862, and several 

 times subsequently), according to which these so called fibres have the 

 nature of canals. While the various authors who use the terms canal or 



