Fluctuating or Alternating Hydration Effects. 



81 



experiments had been extended. It is to be noted that, in addition to 

 hydration and possible salt formation, the colloid was also undergoing 

 some alteration by the unequal solution out of the solution of pro- 

 tein and agar. 



iZp.m 



I2p.m. m. 



I2p.m. m. 



12 p.m. 



Fig. 12. 

 Continuation of record 

 of variations of sec- 

 tions of agar and bean 

 protein in fig. 11. For 

 description see text, 

 pages 81 and 82. 



IZp.m. 



12 p.m. 



12 p.m. 



12 p.m. 



New sheets were fitted to the recorder of the auxograph and arrange- 

 ments made to follow further changes (see fig. 12). During the next 

 4 days an additional swelling of 280 per cent in the malic series, 320 

 per cent in the citric, and 300 per cent in the formic were recorded. 

 Replacement of the acid by hydroxid (1) resulted first in an expan- 

 sion which was partially lost in 7 hours, so that the net gain was very 

 light. When the hydroxid was washed off (2) hydration in distilled 

 water was followed by expansion of lesser amplitude than in the 

 previous procedure of this kind, but it had not ceased at the end 

 of 14 hours. A diminution in each repetition was found. Hydroxid 

 (3) failed to bring the sections back to the dimensions preceding the 

 last hydration. Replacement of the hydroxid by acid (4) caused a 

 further slight contraction, but not to the last pre-hydration dimen- 

 sions. In fact, every hydration included an irreversible element. 

 Hydroxid (5) again produced shrinkage, and then contraction which 

 soon ceased. After 13 hours in hydroxid, water applied and re- 

 newed (6) produced a swelling which was in progress at the end of 

 12 hours. Replacement with acids (7) was followed by very abrupt 

 shrinkages which were more gradual in formic acid. Substitution 

 of hydroxid after 11 hours at 8 was followed by the expected initial 

 expansion and subsequent shrinkage. The final hydration (9) on the 

 tenth day of the test gave swellings with net expansions of 7, 9, and 

 8, as compared with 21, 18, and 17 on the sixth day. The biocoUoid 

 is thus seen to progress through a period of reactions of increasing 

 amplitude to a climax, followed by one of diminishing alterations in 



