4 U. S. BUREAU OF FISHERIES. 



even 8 hours 15 minutes. Most of the unsegmented eggs at the 

 latter stages on being examined cytologieally appeared to be normal. 

 They were therefore merely lagging behind. The proportion of 

 eggs which has failed to cleave becomes progressively less with age. 

 which in itself supports the idea that we are dealing here rather with 

 a delay in cleavage than with a lack of impregnation. The curve 

 in figure 1 (p. 3) shows graphically the numei'ical conditions en- 

 countered. These data are emphasized here merely to show that it is 

 manifestly impossible even after eight hours to designate eggs as un- 

 fertilized when the absence of segmentation is taken as a criterion. 



Figs. 2 to 10. — Siu-faco views of pike-perch eggs. Magnification, approximately X 100. 



Fig. 2. — Normal 6-hour 4-celled blastoderm. 

 Fig. 3. — Normal 10-liour blastoderm. 

 Fig. 4. — Abnormal S-hour blastoderm. 

 Fii;. ."). — Abnormal lO-hoiir blastoderm. 

 Fig. 6. — Abnormal 10-hour bla.stoderm. 



Fig. 7. — Abnormal 10-hour blastoderm. 

 Fig. 8. — Abnormal 19-hour blas.toderm. 

 Fig. 9. — Abnoiinal 29-hour blastoderm. 

 Fig. 10. — Abmormal 29-hour blastoderm. 



In addition to all this it must be considered that in the artificial 

 insemination of the pike perch the eggs are immersed in milt Avhich 

 is diluted very little. The chance of a normal ripe egg remaining 

 unfertilized must therefore be extremely small, and, as a matter of 

 fact, it is surprising that polyspermy is not more often encountered. 

 In addition to the eggs which are found to be dead almost imme- 

 diately, and to those which are slow to cleave, there is a third class 

 Avhich has been designated as " abnormal." In explanation it must 

 be stated that minor irregularities in cleavage are not necessarily an 

 indication of pathological conditions (H. V. Wilson, 1891), and only 

 such extreme cases as are shown in figures 4 to 10 were rated as 



