I'KKSinVATKU AUU-:. 17 



production of packets for purposes of propagation is not confined to the margin of the 

 thallus, as Wille (Nyt Mag. f. Naturvidenakab, xl., 1902, p. 215) describes it, but 

 can take place anywhere. 



In the above-mentioned paper of Wille's a number of reproductive stages of 

 another kind are described and figured (<>j>. <-it., p. 21G, Tub. III., figs. 13-19) ; these 

 are supposed tu lie the products of unicellular akinetes produced from the margins 

 of the thallus (of. Gay, Recherches sur le developpemeut et la classification de 

 quelques Algues vertes, Paris, 1891, fig. 130). Stages like those shown in Wille's 

 figs. 13-15 were common not only in the material from Cape Adare, but more or less 

 isolated among other Algae from several other localities (notably from the Gap pond). 

 There can be no doubt that these stages arise from large unicellular akinetes, as such 

 were frequently observed in the mature thalli (text-figure A, k}. These akinetes arc 

 however by no means confined to the margin of the thalli, but can arise anywhere. 

 They are generally more or less separated from one another, the surrounding cells 

 being often quite irregularly arranged, pale in colour and apparently in a moribund 

 condition (text-figure A). The akinetes are spherical and vary in diameter between 

 10 and 14/u,; they have a thick membrane. Division of the contents of the akinete 

 (text-figures C and D) does not appear to take place until it is liberated by the decay 

 of the surrounding thallus, but there may be exceptions to this rule. 



I have no doubt that Wille is right in regarding these cells as sporangial, as the 

 irregularity of arrangement of the products of their division excludes the possibility 

 of their being the beginnings of a new thallus. Wille thinks that the product's 

 (his aplanospores) on liberation give rise to the fformidium-st&ge. While this may 

 frequently be the case, there seems to be indirect evidence that they may sometimes 

 give rise straight away to the characteristic 7 'n'.v/W,/ -packets. The abundance of such 

 packets on the Phormidium-sheete from many localities is scarcely explicable on any 

 other hypothesis. There are no mature Pi'asiola-t\i&\\i, from which these packets 

 could have been derived, at hand in these localities, whereas the dividing akinetes 

 are not uncommon. 



The filaments of the Homin/it/iti-stage from Cape Adare measured 9-1 1 p in 

 diameter. 



12. PRASIOLA ANTARCTICA. 



(Text-figures E and F.) 



Prasiola antarctica Kiitz., Spec. Alg. (1849), p. 473 ; Tabul. Phycol., v., Tab. 4<i, fig. t ; Rabenliorst, op. 



</!., iii. (IM;S), p. 311. 



Hob. Growing on stones, Mt. Terror, January 22nd, 1902. 



It is with some degree of diffidence that I refer the form from Mt. Terror 

 to a species distinct from P. crispa, as I do not feel quite convinced that the 

 differences are of specific value. P. <.<nt<<r<'ii<-<i might perhaps be regarded merely 

 as a variety of P. cris^a, as 1 advocated in my paper mi the Algae of the 



i i- 



