104 E. A. ANDREWS. 



8) Branchlostomum pdarjicum, 36? 10? 15-67. 10 mm. 



9) Asijmmetron Jucayanmn, 44. 9. 13-06. 13 ,, 



In this .table the first cohimii of figures indicates the number of muscle 

 segments anterior to tlie atriopore ; the second the number between the atriopore 

 and the anus ; the third the number posterior to the anus, that is in the tail. 

 These three columns thus serve to indicate the relative length of the three 

 main divisions of the body as well as to state the actual number of segments 

 in each ; from them we may judge what the general aspect of a species will 

 be, whether with a short tail, a long trunk or with closely approximated 

 atrial and anal openings etc. The fourth column represents the average 

 number in the entire animal ; it is not always the sum of the preceding 

 three clumns though it would be so in any particular aminal. The last 

 column gives the entire length, on the average. 



The six lancelets from Japan have each sixty-four segments, of which 

 thirty-seven are anterior to the atriopore, sixteen between the atriopore and 

 the anus and eleven posterior to the anus. These numbers were found to be 

 the same on the right and on the left in each case. The length of the speci- 

 mens was 85.5 mm. 45 mm. 29.5 mm. 45 mm. 31 mm. and 40 mm. in the 

 six specimens : that is an average of 39 mm. nearly. As far as can be 

 judged from these specimens the formula for the species would thus be : 37. 

 16. 11-64, 39 mm. 



This we see does not agree with that of any of the known forms, but we 

 need not conclude that it is new on that account since the formulas are but 

 approximately acculate. They are averages, often from only a very few 

 specimens and the actual numbers may vary considerably in a given species- 

 Thus in B. cuUeUum WiWef counted 32. 12. 8-52 and 33. 11. 11-55 and 

 again in some cases a total of 54. Then again the errors in counting are 

 large owing to the difficulty of locating the atriopore and anus accurately and 

 to the difficulty in seeing the boundaries of the last segments. 



The entire number of segments in the Japanese specimens is almost the 

 same as in the formula for B. Bekheri while the number in the anterior is 

 thirty-seven in ease case : there is thus a closer agreement here than with 

 1 Quart. Journ, Micr. Sc. January 1894.. 



