34 W. WOODLAND. 



space disposition of the scleroblasts coucerned in their indivi- 

 dual formation : the elongated needle is deposited by a bi- 

 nucleated mass of scleroplasm with maximum and minimum 

 diameters in the plane of the body-wall, and tlie spherical 

 granule by a scleroblast whose corresponding diameters 

 are all approximately equal. The granule in A. elegans 

 next assumes a flat three-cornered shape whilst still con- 

 tained by the single scleroblast, and the three corners 

 of the triangle thus formed then elongate to form a young 

 triradiate spicule (fig. 3). Shortly after this stage is reached 

 the nucleus of the scleroblast divides and two scleroblasts in 

 consequence appear in connection with the young triradiate 

 (fig. 4). Approaching nuclear division in the scleroblast is 

 always denoted by large size and faint coloration by picro- 

 carmiue (tig. 5a, e.g.) ; as is well known picro-carmine is a 

 stain which never rendej-s the details of karyokinesis visible. 

 The three arms of the spicule next show signs of bifurcation 

 at the extremities (figs. 5, 6), iind, indeed, the whole of the 

 further development of the spicule (when this is not situated 

 at the extreme edges or ends of the arms, in which position 

 certain parts elongate greatly in a distal direction) consists, 

 as in Cucumariidse, of a series of bifurcations resulting in a 

 more or less circular perforated plate (fig. 7), but it is notice- 

 able that the attached scleroblasts differ from those of Cucu- 

 mariidse in their much greater number. The fact that 

 abnormally-large nuclei are so often met with in these 

 scleroblasts renders it exceedingly probable that they are all 

 (thirty or forty in the case of the larger plates) derived from 

 the original mother-scleroblast, although it is evidently im- 

 possible to make a decisive statement to that effect. This 

 repeated division of the mother-scleroblast might, indeed, be 

 attributed to the necessity for the proximity of nuclear sub- 

 stance to the constantly-increasing area of deposition (the 

 well-known experiments of Verworn on Polystomella proving 

 enucleated portions of protoplasm to be incapable of secreting 

 a shell pointing the argument) were it not that no such sub- 

 division occurs in the case of other spicules of equal size 



