— 96 — 



rosettes, sheaves or star-shapcd masses which remine! one of t)'- 

 rosine crystals, as will be seen in figures 4 and 5. 



Fig. 5. — Haemocyanin-crystals from the diluted blood of Octopns viilgaris 



It is difficult to define the conditions necessary to obtain 

 " projectiles " or needles of haemocyanin. 



On employng the Kopkins-Pinkus niethod (10) for crystalh- 

 zation 1 obtained " projectiles " more frequently, and on emplo- 

 ying the unniodified method of Hofmeister (9), needles were mo- 

 re frequently obtained, but this is no rule. Perhaps the increa- 

 sed acidity of the solution, in the Hopkins-Pinkus modification, 

 may determine the formation of " projectiles "; this however cannot 

 be affirmed. Very often also, in this case, the haemocyanin cry- 

 stailizes in needles. 



Redissolving the needles with a little water, and allowing the 

 solution to recrystallize, " projectiles " are very often obtained. 

 Sometimes I obtained projectiles of haemocyanin only after ha- 

 ving recrystallized the needles several times. 



Not unfrequently 1 obtained from the sanie solution a mi- 

 xture of needles and " projectiles ", the needles being the first to 

 form. The liydrogen-ion concentration of the solution does not * 

 seem to have mudi influence on the crystalline forni. 



