June 7, 1907] 



SCIENCE 



913 



freely interconnected. The sponge in this 

 condition closely resembles Spongilla in its 

 ■winter phase, as described by Weltner.' Pre- 

 sumably water continues to circulate through 

 the body, but the current must be an exceed- 

 ingly feeble and irregular one. 



As a sponge in this condition continues to 

 shrink, it may subdivide and thus a large 

 sponge may eventually be represented by 

 numerous masses, in a given case about 1 cm. 

 in diameter. Now if the sponge in this con- 

 dition or if one of the masses into which it 

 has split up, be attached to wire gauze and 

 suspended in a live box floating at the sur- 

 face of the open water of the harbor, the 

 sponge or piece will in a few days grow and 

 redevelop the pores and oscula, flagellated 

 chambers, tissue differentiation, and skeletal 

 arrangement of the normal sponge. Whether 

 in this regeneration the transformed and 

 separated collar cells again unite to form the 

 flagellated chambers, I can not say. I think 

 it very doubtful. 



In the two classes of cases just described 

 the sponge as a whole degenerates and slowly 

 shrinks. Cellular death takes place so gradu- 

 ally that at no time is there any obvious 

 corpse tissue or skeletal debris. Much more 

 common and of far greater interest are the 

 following cases. In these a large part of the 

 sponge body dies in the course of two or three 

 weeks, leaving the skeletal network still in 

 place and bearing the brown decaying rem- 

 nants of the flesh, which, as maceration con- 

 tinues, are washed away. In places, however, 

 the sponge body does not die. Here masses 

 of living tissue are left, conspicuous amidst 

 the dead remains by their bright color and 

 smooth, clean surface. These living frag- 

 ments may be classified into three groups. 

 First, the upper end of an ascending lobe or 

 a considerable part of the body of the lob« 

 may be left alive in its entirety, thus forming 

 a more or less cylindrical mass up to 5 mm. 

 diameter, with a length sometimes two or three 

 times the thickness. The histological condi- 

 tion of these masses is not very different from 

 that of the sponges already described. Such 



' ' Spongillidenstudien, II. Archiv fur Natur- 

 geschichte,' 1893. 



a mass may be said to consist of anastomosing 

 trabeculee, separated by the remains of the 

 canal system. The mesenchyme composing 

 the trabecule consists of discrete cells inter- 

 connected by processes to form a syncytium. 

 The flagellated chambers as such have nearly 

 disappeared, although remnants may still ba 

 recognized. In them the collar cells have 

 transformed into simple polyhedral bodies that 

 are widely separated. The bulk of the cham- 

 bers have broken up into their constituent 

 cells, and these are now scattered as ele- 

 mentary parts of the general mesenchyme. 

 When such masses are attached to wire gauze 

 and hung in a floating live-box they trans- 

 form into perfect sponges. 



A second class of surviving remnants in- 

 cludes masses scattered over the general sur- 

 face of the sponge. These may be spheroidal 

 and small, less than one millimeter in di- 

 ameter. Usually they are flattened and of an 

 irregular shape with lobes, suggesting a lobose 

 rhizopod or myxomycete Plasmodium. Such 

 masses which may be connected by slender 

 strands are commonly from two to five milli- 

 meters in the longest direction. The third 

 class of remnants are found scattered through 

 the body of the dead and macerated sponge, 

 in which they sometimes occupy positions that 

 are obviously favorable for respiration. These 

 bodies are more or less spheroidal and small, 

 their diameter varying commonly from one 

 half to one and a half millimeters. In the 

 most successful cases of treatment, the small 

 masses, internal and superficial, are exceed- 

 ingly abundant, and the dead and macerated 

 sponge body with its contained nodules of 

 conspicuous living tissue strongly suggests a 

 Spongilla full of gemmules. 



These living remnants of the sponge (bodies 

 of the second and third classes) execute slow 

 amoeboid changes of shape and position, be- 

 having thus like plasmodia, and they may be 

 designated as plasmodial masses. Micro- 

 scopic examination shows them to be of an 

 exceedingly simple character, without canal 

 spaces or flagellated chambers. The mass does 

 not consist of discrete cells, but is an aggre- 

 gation of syncytial protoplasm studded with 

 nuclei. The protoplasm is stored with minute 



