Genital System of the Myxinoidea 95 



and August, 1899; the double line in the lower left corner represents the percentages of 

 females caught during the same months. 



The heavy broken line (short dashes) represents the percentages of males of H. 

 burgeri taken in the various catches made in Japan from July 18, 1900, to July 25, 1901. 

 The heavy solid line represents the percentages of all females to the total catches of H. 

 burgeri made in the various months of 1900 and 1901. 



The thin solid line represents the percentages of females of H. burgeri which con- 

 tained eggs of approximately uniform si2;es, taken in the various catches from July 18, 

 1900, to July 25, 1901, as shown in column ten of Table II (p. 81). The thin broken 

 line (long dashes) represents the predominating si2;es of eggs in females of H. burgeri 

 taken in the various catches throughout the year from July 18, 1900, to July 25, 1901. 

 The ordinate in this case represents the length of eggs in millimeters, written in arabic 

 numerals. 



Between the dates of August 29 and October 28, 1900, Dean was unable to catch any 

 eels, although the lines were set several times. He believed that his failure to get any 

 specimens during this time was due to the fact that the eels were breeding; it is well 

 known that many different kinds of fish do not feed during the spawning season. 



Further evidence that the spawning season may be during the months indicated by 

 Dean in his graph is furnished by the fact that, immediately after the spawning season, a 

 very high percentage of the females taken had empty ovarian follicles (capsules), as is 

 shown in the graph. Table II (page 81) shows the percentages of females having empty 

 follicles in the ovaries to be as follows : on October 28, 1900, 63 per cent; on October 30th, 

 90 per cent; on November 25, 98 per cent. The presence of empty ovarian follicles in the 

 ovary of a female is positive proof that ova had been recently deposited. 



Additional support for Dean's opinion as to the breeding season is to be found in his 

 data regarding the si^es of eggs in females taken in the various months of the year. In 

 Table II (page 81) it will be seen that on December 15, 1900, one month after the catch 

 which contained the highest percentage of empty ovarian follicles, the average length of 

 the eggs in the females was 4.4 mm.; on January 15, 1901, one month later, the average 

 length of the eggs was 4.2 mm., or practically the same. This lack of increase in size can 

 be explained by the hypothesis which I advanced (1917, P- 124), namely, that the eggs in 

 an ovary containing corpora lutea are arrested in development until the latter are almost 

 completely resorbed. On March 30th, two and one-half months later, the average length 

 of the eggs was 11.9 mm. From that date on there was a gradual increase in the size of the 

 eggs, the largest ones being in females caught on August 29th, shortly before the spawning 

 season. These eggs in females taken on that date averaged 23 mm. in length, which is 

 approximately the size of mature eggs. These facts in regard to the gradual increase in the 

 size of eggs from December to late August are shown in the graph (page 94) by means of 

 the long dashes. In this case the abscissa, as before, represents the months of the year, 

 while the ordinate indicates the average length of eggs in millimeters. 



