368 



nised by outward appearances. Nevertheless, an anatomical examination is 

 necessary to understand its character. Its structure consists of fibres, 

 somewhat symmetrically arranged to the long axis of the branches, connected 

 by secondary fibres at right angles, making thus very defined channels. 

 But this would not be sufficient to discriminate it, inasmuch as the next 

 genus I shall speak of is similar, and, indeed, is very closely allied. Its 

 name is " Chalina " (anchor), and its intimate structure shows us a 

 horny fibre, within which spicules are developed in lines. In C. oculata, 

 that which I refer to, the spicules are singly thus arranged, but in others 

 they are more numerous. It sometimes shows the habit of taking up 

 extraneous particles, such as spines of small echini, &c. I thus pass 

 away from the horny sponge in these examples, which show a connec- 

 tion with the order, " Silicea," which connection in some foreign sponges 

 is extremely curious, and much more developed. 



Isodictya comes next; the name implies a similar or formal network. But, 

 in this arrangement, it shares in some degree with the last genus. From 

 this, however, it differs through the absence of keratose fibre, the spicular 

 network being often formed of single spicules, attached at their terminal 

 points by keratode. /. simulans being somewhat branched, allies itself in 

 form with the last ; so also do some species where the keratose matter on 

 which the spicules of the skeleton lay, but not imbedded within it, is much 

 developed. Indeed, the alliances of genera are often thus shown, "where 

 there is a departure from one type, at the beginning, and, at the end, 

 where it may overlap or ally itself with another. In many of the mem- 

 branes of this genus, the network is made up of single spicules in union 

 with each other, and in the fibre the spicules may be so numerous as to 

 simulate that of Desmacidon. Of all Dr. Bowerbank's genera this has 

 the greatest number of species, viz., 70, and it naturally embraces the 

 sponge in all its forms and characters. It is massive, or coating, para- 

 sitic, or self-contained. It is branched or amorphous, cup-like and fan- 

 like, and in its natural condition of all kinds of colours. It abounds in 

 almost every variety of spicule except the stellate, or forcipiform. It 

 is sometimes very minute, and of British sponges it claims to have the 

 largest in J. palmata. With the structure of this genus may be natur- 

 ally associated the two British freshwater sponges ; these, however, form 

 a class by themselves, and have had special notices in our Journal. 



I will now treat of the genus Halichondria (sea-bread), in which 

 alliance may be found in the membranes of some of the Hymeniacidon, 

 because of their reticulated structure, and some resemblances may also 

 be seen in the last-described genus. Halichondria , however, as defined 

 by Dr. Bowerbank, has a definitely constructed reticulation through the 

 entire sponge, not confined to a particular part. The rete is multi-spicular, 

 held together at the terminal portions of the spicule, and is generally well- 

 defined. Two of this genus are exceedingly common, found almost on 

 every shore, and both particularly characteristic. One is H. panicea, the 

 bread-crumb of Ellis and of our early naturalists, and this is easily dis- 

 covered by a hand magnifier, which will show the marked distinction of the 



