388 DISCOVERY REPORTS 



black such as occurs in the arthropod cuticle; however, the value of the latter stain in the histo- 

 chemical identification of chitin is doubtful. 



The main value of the silver preparations has been to show how the cuticle is attached to the 

 underlying epithelium. In certain of the silver preparations of whole strips of codon-ectoderm, the 

 cuticle shows up well as a dark sheet of granular material, split in places and flaking away, revealing 

 the ectoderm underneath. Around the edges where the cuticle is damaged, one can see dark strands 

 of cuticular substance passing down into the ectoderm. In places where the cuticle has disintegrated 

 more completely, these strands are often all that remains of the cuticle. They are funnel-like structures, 

 rooted in the epithelium by means of a basal swelling. They may run almost perpendicularly from the 

 cuticle proper down into the ectoderm, but more often they run obliquely. In such cases, they can be 

 seen to good advantage in surface view, that is to say in the strip-preparations. PL XXVII, fig. 5 shows 

 two of these funnel-like cuticular tubes running down into the ectoderm, where their swollen basal 

 parts are rooted. The details of the ectoderm cells are not shown in this photograph, as the silver has 

 precipitated heavily and coarsely over the whole exposed surface. The basal swellings are light in colour, 

 and appear to be hollow. Sometimes the stalk or funnel region is also hollow in appearance. 



These structures appear to consist of solidified streams of cuticular substance running from extinct 

 gland cells (represented in outline by the swollen bases) up into the cuticular sheet. This was also 

 von Lendenfeld's interpretation of a closely similar histological picture in Eucopella (Lendenfeld, 

 1883). Apart from these structures, no connections between the cuticle and ectoderm have been 

 found. 



In the case of the pneumatocyst, or cuticle of the saccus, the means of attachment to the ectoderm 

 again appears to consist solely of swollen portions of the cuticle embedded in the ectoderm, and again 

 it is probably fair to say that these represent the sites of extinct gland-cells. The globular thickenings 

 are black in unstained preparations, and can be seen without difficulty. They proceed directly from 

 the underside of the cuticular sheet, and there is no stem or funnel region as in the codon. The only 

 part of the saccus not lined by the pneumatocyst is the gas-gland. This becomes explicable if it is 

 assumed that the pneumatocyst has the properties of an insulator against gaseous diffusion; in the 

 main part of the saccus its function would be to prevent loss of gases by diffusion, but in the region 

 of the gas-gland, where gases enter the float-chamber, its absence is a functional necessity. Certain 

 other findings support the view that the pneumatocyst may function as an insulator (see page 389). 



The chemical composition of the cuticle or pneumatocyst in the Siphonophora has never been 

 satisfactorily determined. In Velella, which is not a true siphonophore (see Totton, 1954), it is clear 

 that the substance lining the concentric float-chambers is chitin (Leuckart, 1852; Henze, 1908; 

 Rudall, 1955), but apart from some unsatisfactory early work by Leuckart on Physalia, nothing has 

 been done on the siphonophores. We know from an earlier paper (1851) that Leuckart thought the 

 saccus to be a homogeneous, ' structureless ' substance, like the chitinous skeleton of Velella, whereas, 

 in fact, it is living tissue, composed of ectoderm, endoderm and mesogloea, with only a thin membrane 

 of what might be and, as we shall see, actually is chitin. Thus when Leuckart (1852) reported that the 

 saccus consisted of chitin, we cannot place much value on his finding. Nearly all the saccus dissolves 

 in caustic potash, as he would surely have noticed if, as he claimed, he used this reagent. 



In the present investigation a modification of Campbell's chitosan method (Richards, 1951) was 

 employed to test for the presence of chitin in the pneumatocyst. A whole saccus was used in one test, 

 and the dissected pneumatocyst in another. 



The tests gave a positive result. It was necessary to shorten the period in caustic alkali to 2 min., as 

 after this the membrane tended to disintegrate (but not to dissolve) at 160 C. This curtailed treatment 

 has a precedent in Richards's method for butterfly wing scales, which are also thin and fragile. In the 



