488 G. N. CALKINS AND L. H. GREGORY 



tests was 48 pairs in J 1, 49 in J 7, 34 in J 8, and, 9 in J 21, all 

 from quadrant A while not one pair was found in the other twenty 

 pure lines. A mixture of J 1, J 8 and J 21 gave more than 500 

 pairs. The result of this test for one date is shown in diagram 

 3, December 5. 



For the January test the rich cultures were seeded on the 26th 

 of December and tests were made at different times, the aggre- 

 gate result being 388 pairs in Jl; 135 pairs in J 7; 528 pairs in 

 J 8; 816 in J 21, all in quadrant A and not one pair in the re- 

 maining twenty lines. 



For the February test the rich cultures were seeded February 

 1 and the tests were made on the 3rd, 5th and 9th, resulting in 

 3 pairs in Jl; 136 in J7; 138 in J8; 39 in J21, again all in quad- 

 rant A, and not one pair in the remaining nineteen lines (no. 

 J 14 died out January 19). 



For the March test the rich cultures were seeded March 6 and 

 the tests were made on the 9th, 13th and 15th. The earlier his- 

 tory was again repeated but larger numbers of pairs were found 

 in the conjugating lines of quadrant A: 890 pairs in J ; 1854 pairs 

 in J 7; 2118 in J 8, and 1400 in J 21. Again not a single pair 

 was found in the remaining nineteen lines. The five monthly 

 tests from permanent culture material which had not conjugated 

 since isolation of the ex-conjugant in July, thus show a remarkable 

 physiological variation in the progeny from that ex-conjugant. 

 The variations furthermore were remarkably consistent during this 

 period, the lines that had conjugated at the first test, continued 

 to give conjugations at each subsequent test and would have 

 given conjugations at any time the test might have been made. 

 During all of this period the non-conjugating lines were equally 

 consistent, not a single pair occurring in all of the non-conjugating 

 lines. The interest of this finding is heightened by the fact that 

 the conjugating lines all come from one quadrant of the ex-con- 

 jugant which was separated from the other three quadrants at 

 the second division after conjugation. 



The April test brought a surprise. The rich cultures were 

 seeded March 31 and tests were made on April 5, 6 and 9, result- 

 ing in epidemics of conjugation in the usual lines, J 1 with 3874 



