DEVELOPMENT OF SPONGES FROM DISSOCIATED TISSUE CELLS. 21 



glass as thin incrustations. The latter contracted and fused with one another, forming 

 massive bodies and loose aggregations of such bodies. Massive collections of tissue of 

 this sort, as has been recorded, metamorphosed in the live box. 



The effort was made to see if such massive collections of sponge tissue would not 

 metamorphose in laboratory dishes. Small massive bodies were selected which for 

 one reason or another had remained quite free, viz, unattached to substratum. These 

 were 500-700 /t in diameter. They had a dense syncytial structure, were homogeneous 

 in appearance, and had a quite smooth surface — in short, were very gemmulelike. 

 They were kept in laboratory dishes with Ulva, but would not transform, behaving then 

 in like manner to the masses of regenerative tissue that form in a degenerating sponge 

 or are produced outside the sponge body. 



Several similar masses were put on July i in Ulva dishes. Some made a slight step 

 toward metamorphosis, in that they flattened out at points of the periphery, here spread- 

 ing for a short distance over the substratum. Regressive changes then occurred, and 

 on July 27 the bodies were again spheroidal and smooth. 



The conclusion seems to be that when the sponge cells have once united into a 

 compact mass of any size, this mass is slow and as it were reluctant to transform. Par- 

 ticularly is this so if the mass of tissue has been free long enough to acquire a smooth 

 surface. It has by this time apparently passed into a quiescent physiological state. 

 For such a mass to set up differentiation, the stimuli coming from the open water (excel- 

 lent aeration and movement of water probably) are necessary. 



Experiment record 7, July 21, igoj. — Question involved: Will compact masses 

 formed by the continued union of tissue teased out of the sponge metamorphose in 

 live box? The masses began but did not complete the metamorphosis. Essentially 

 same experiment as 4. 



Fresh Mwrociona tissue was teased up and the teased-out tissue allowed to fuse. 

 The bulk of the tissue was collected in center of dish, where numerous compact masses 

 commonly 0.5 to i mm. in diameter were formed. Most of these were more or less 

 united to form larger aggregates. 



Two of the compact masses were hung in bolting-cloth bags in the live box on 

 July 22. One was smooth, spheroidal, 800/i in diameter. The other was a composite 

 mass of same character as that shown in figure 27, about 3 mm. long, 2 mm. wide, and 

 0.5 mm. thick. It included a small fragment or two of the old homy skeletal frame- 

 work. On July 29 the smaller mass had split into two, each of which was a flattened 

 incrustation firmly attached to the bolting cloth. The two incrustations were oppo- 

 site, and it was evident that the original mass had attached to both surfaces of the bag. 

 The larger mass had likewise split into two, both of which had flattened out and attached 

 to the cloth. These masses went no further in metamorphosis, but eventually died. 



Experiment record 8, July 30, 190J. — Result: Teased-out tissue quickly combines 

 to form small masses. These fuse if brought in contact. If not too large they then 

 flatten in peripheral region which spreads over substratum. On same day tissue pressed 

 out through bolting cloth was found to behave in same way. 



