THE COLLAR-CELLS OF HETEROCGLA. 33 
mettre en suspension des granules de carmin, provoque l’occlu- 
sion relativement rapide des papilles.” It is not clear from 
the words whether this is due to the presence of particles or 
only to stirring the water; but it is well known that these 
papille are exceedingly sensitive. For other sponges, and 
especially S. raphanus, Lendenfeld makes repeated state- 
ments (11, pp. 588, 592, 675, &c.) as to closure of pores 
against carmine (and not against milk). They are contradicted 
by the experiments of every other worker; and notwithstanding 
their picturesque elaboration, and the dramatic deductions for 
which these statements are responsible, the 149 experiments 
that he records include no evidence that the narrative is based 
on even erroneous observation. 
DistorRTION OF CELLS IN PRESERVATION. 
The following results may be of some interest to those who 
study histology on preserved material from other groups as 
well as sponges, though the measurements are too few to 
profess to be more than suggestive. 
Measurements were made of the collar-cells in six series of 
sections, A, B, C, D, E, F, in order to compare their dimen- 
sions with those of life. The series were from five specimens 
(S. compressum), D and E being from one sponge; and in 
the case of all but A the collar-cells from a closely adjoining 
portion of the same individual were examined and measured 
during life. 
All the sponges were preserved in osmic acid for one hour ; 
followed by alcohol, benzol, and paraffin. In C, D; and F the 
change from water into absolute alcohol was effected by dia- 
lysis; in all but B the change from absolute alcohol into 
benzol was made in the same way; all were transferred by 
gradual changes of temperature and percentage through soft 
paraffin to hard paraffin of a temperature not exceeding 65°C., 
generally 62° C. 
A was the only sponge preserved in + per cent. instead of 
ai per cent. osmic acid, it alone was decalcified (1 per cent. 
nitric acid in 90 per cent. alcohol), it alone was stained in 
voL. 38, PART 1.—NEW SER. C 
