1914 ] N. AnnandaIvE : Sponges of Lake Baikal. 143 



channels never running in a horizontal direction immediately 

 below the dermal membrane ; inhalent subdermal cavity absent. 



Skeleton consisting of a network of well-defined, compact, 

 strongly coherent series of spicules lying parallel or nearly parallel 

 to one another in a thick sheath of chitinoid substance. The vertical 

 fibres branch dichotomously , especially in the outer part of the 

 sponge, and are joined together by transverse fibres containing 

 fewerspicules than themselves. On the surface branching becomes 

 more vigorous and more irregular, so that the external extremi- 

 ties of the vertical fibres form broom like bunches of slender 

 fibres the central part of which is as a rule hollow and forms a 

 nursery for the young embryos. Together these bunches of 

 fine vertical fibres constitute a skeletal cortex (pi, ix, figs, i, la). 



Spicules. — There are no true microscleres. The skeleton 

 spicules are aniphioxous and spiny, the spines being sometimes 

 concentrated at or near the extremities. Smooth slender amphioxi 

 also occur occasionally in the parenchyma. 



Type species : Spongia baicalensis , Pallas. 



No gemmules have been described in this genus and the form 

 of the frei-swimming larva is unknown. Embr3'os in an early 

 stage of development are frequently present in large numbers ; 

 they appear to migrate to the cavities in the terminal bunches of 

 the spicule-fibres and probably escape thence on reaching the 

 larval stage. 



The only species that can be assigned to the genus are L. hai- 

 calensis (Pallas) and L. abietina (Swartschevski). The latter has 

 been found only in Lake Baikal, but the former occurs also in 

 Arctic seas.' 



Dybowski has described (1880) several well-defined varieties 

 of L. haicalensis. but Soukatschoff's* L. baicalensis var. e cannot 

 belong either to the species or the genus. It is probably a 

 form of Baikalospongia bacillifera (Dybowski). A phase to 

 which no name or letter has been assigned was submitted to me 

 by the authorities of the St. Petersburg Academy. In it the 

 upright part of the sponge, instead of consisting as in the typical 

 form of cylindrical systems, is fan-like, the broad, compressed 

 growths usually being curved in horizontal section and sometimes 

 forming incomplete cups. This form evidently reaches a consider- 

 able size. Its spicules agree with those of the typical form. 



L. baicalensis and L. abietina differ mainly in the structure of 

 the skeleton ; in the latter the vertical fibres branch much less 

 freely, the skeletal cortex is less well developed and the transverse 

 fibres are fewer and more slender than in the former. L. abietina 

 never produces upright growths like those characteristic of the 

 typical form or the phase described above of L. baicalensis, but 

 the formation of such growths does not take place in all varieties 

 of the latter species. 



i Dybowski, Sitzb. Nat. Gesellsch. Dor pat, 1884, p. .14. 

 '^ T'rav. Soc. Nat. St. Petersb. XX\', p. 11 (1895]. 



