Discussion. 



Horse SH 1. — The haemolytic index was higher for mules than for 

 hoi^ses. In no instance was an injection of this serum followed by 

 haemolysis. Horse serum with a haemolytic index of 0"3 for liorses did 

 not seem to have any harmful effect on horse lilood in vivo. 



Horse 147. — The serum was in one instance more haemolytic foj' 

 mvdes than for horses : in another test vice versa. When injected it 

 killed four horses out of five. 



Horse 160. — This serum seems to have been more haemolytic for 

 mule than for horse blood. With an index of 1'3 it did not prove to 

 be haemolytic in vivo. (All animals being injected subcutaneously.) 

 With D'egard to the rise and disappearance of the isolysines, it may be 

 observed that from the 29th September, 1903, to 2oth April, 1904 (seven 

 months), its haemolytic index for horses fell from 1"3 to 0"284, although 

 two injections of blood, totalling 2,500 c.c, had been made. We must 

 consider, therefore, that within the thi'ee months which passed from 

 the date of haemolysis test, 29th September, 1903, to the tenth injection, 

 12tli .January, 1904, a large amount of the isolysines — perhaps all — had 

 disappeared. I am not able to say that the loss of 32,880 c.c. of blood 

 is the cause of the loss of haemolysines ; the haemolytic index for mules 

 had, however, increased between 25th April to 26th September, 1904, 

 from O'o to 0'84, notwithstanding the withdrawal of 35,000 c.c. blood 

 from the horse, and that no further hyperimmunisation had taken 

 place. 



Horse 172. — The haemolj^tic index was higher for mules than for 

 horses. 



Horse 611. — This serum also had a stronger haemolytic power for 

 mule than for horse blood. The haemolytic indices were somewhat in- 

 creg^sing from the third to the sixth injection ; injections of 4,500 c.c. 

 within three months did not give rise to haemolysines for horses. 



Horse 612. — This serum was also more haemolytic for mules than for 

 horses. The activity for horses — adversely to h. 611 — decreased, although 

 three more injections were made. Both liorses 611 and 612 were injected 

 with eflual quantities qf blood of the same origin at the ^ame time, and 

 the haemolysis tests were made on t|:^e same days ; but the quantities 

 of haeraolysines which they deyeloppd in tlieir blopd We^-e different. 

 For inst3,nce, the haemolytic index for mule ^ood 611=0'2 fpr 612 = 0'125 

 at the same elate. The only difference between the horses at this date 

 is that 612 had been bled i/O the extent of 6,000 c.c. more than 611. (Com- 

 pare h. 160.) 



Whilst the haemolytic index in 611 increased during three months, 

 it decreased in 612. The complete disappearance of haemolysines in 612 

 within two months is demonstrated by the haemolytic index (for horses) 

 of 0167 on the 29th November, 1904, and nil on the 2nd February, 

 1905. These horses clearly show by their paraUel treatment the indi- 

 vidual variations and differences in the appearance, disappearance, and 

 absolute efficacy of haemolysines. 



