572 STUDIES ON SPONGES, 



then been brought forward seems to indicate that the silicious 

 sponges are the ancestors of the horny ones, a view advocated 

 by Vosmeer and Polejaefi'. 



I think that the Chalinojisinas are very nearly allied to the 

 Chalinidse, but I do not think that they are so nearly related to 

 each other as the similarity of their outer appearance would 

 indicate. It seems most probable that the two species of Chalinopsis 

 described above are descendants of digitate Chalinidae, they have 

 lost the defensive spicules which are no doubt of great value to 

 Dactylochalina, but they have retained the outer appearance. It 

 is probable that the Dactylochalina species have undergone changes 

 since then, and that these species of Chalinopsis have had to change 

 their own shape accordingly so as always to remain similar to 

 a defensive sponge. I would therefore call the similarity in these 

 two cases, although it has originated in true relationship, Mimicry, 

 because the structure of the important internal organs has changed, 

 ■whilst no difference is perceptible in the outer appearance -VYhich is 

 so very variable in sponges. 



EXPLANATION OF PLATES. 



Plate, .39. 



Fig. L — DendriUa cavernosa. R. v. L. 

 Section through the sponge. 

 Jth of the natural size. 



Fig. 2. — DendriUa cava-nona. R. v. L. 



Pore membrane covering the vestibule lacunai. Seen from the 



surface. 

 1 :20 magnified. 

 Alcohol, Htematoxylin specimen. 



Fig. 3. — DendriUa cavernosa. R. v. L. 



Pore membrane covering the vestibule lacuna;. Seen from above. 



Margin of one of the pores. 

 1:20 magnified. 

 Alcohol, Hiematoxylin specimen. 



