l8o NATURE AND AFFINITIES OF THE SPONGES. 



aperture or blastopore. The question at issue is whether similar or equi- 

 valent developmental steps are traceable in the ciliated sponge-gemmule ? 

 F. E. Schulze, writing of Sycandra raphanus in the year 1875, deposes that 

 there are, giving in demonstration the figures reproduced at PI. IX. Fig. 33. 

 The same authority reports, however, as the result of a more recent investi- 

 gation of this species, an entirely opposite plan of structure. According to 

 his later interpretation,* it is not the larger and apparent endodermic blas- 

 tomeres that become invaginated or enclosed within the more minute 

 ectodermal elements, but, as indicated at Fig. 34 of the same plate, the 

 latter that sink down into, and are enclosed by, the former. By Metschni- 

 koff, a second chronicler of the developmental phases of this identical species, 

 the so-called ciliated larvae are described as presenting, in addition to the 

 ordinary form having dissimilar hemispheres of large, subspheroidal, non- 

 flagelliferous, and more minute columnar flagellate cells, examples that are 

 made up entirely of flagellate columnar elements, which, however, towards 

 the basal region, are of somewhat larger size. This slight modification of 

 the first of the three structural types enumerated in a preceding page, as 

 observed by Metschnikoff, of Sycandra raphanus, represents the dominant 

 form found in other closely allied calcareous species, as also in the 

 Myxospongiae and the majority of the Siliceospongiae, where the charac- 

 teristic amphiblastuloid type previously considered is not known to occur. 

 This so to say homoblastic embryonic form, produced by the entire and even 

 segmentation of the primitive so-called ovum, and exhibiting in its charac- 

 teristic state the structure and condition only of a monoblastic or simple 

 single-cell-walled morula, with a more or less extensive segmentation cavity, 

 does not subsequently, by direct or indirect invagination, as occurs with the 

 monoblastic morulae of the Metazoa, attain to the higher diploblastic formula ; 

 it cannot therefore be consistently compared with the typical and charac- 

 teristic diploblastic embryo of any Metazoic organism. Nevertheless, various 

 more or less arduous attempts have been made to demonstrate that even 

 in this simpler monoblastic reproductive body, the essential Metazoic 

 elements, " ectoderm," " endoderm," and in some cases even " mesoderm," 

 are substantially represented. 



By some, including Professor Haeckel, it has been suggested that the 

 endoderm is indirectly produced through the process of delamination, 

 instead of that of invagination, as most usually obtains. Such an interpre- 

 tation, however, is entirely upset by some highly remarkable results of 

 recent investigation. It has been shown, in fact, by Oscar Schmidt, f with 

 relation to the calcareous type Ascetta primordialis, that the elements usually 

 accepted as representing the endoderm are produced neither by delamina- 

 tion nor by invagination, but creep into the central segmentation cavity 

 as separate and independent amcebiform units from the circumjacent 

 so-called ectodermal layer, of which latter they are the metamorphosed 



* 'Zeitschrift fur Wissenschaftliche Zoologie, ' Bel. xxxi., 1878. 

 t ' Archiv fiir Mikroskopische Anatomic,' Bd. xiv., 1877. 



