894 OBSERVATIONS ON PLANT PLANKTON. 



by the thickness of the girdle membrane. When a minimum size 

 is reached, diatoms become re-established at the maximum by the 

 formation (in various ways) of auxospores. Mr. Comber has called 

 attention (Jour. Roij. Micr. Sue, October, 189G) to the formation in 

 certain diatoms of " Endocysts " resulting in the production of a 

 dimorphic form, and Prof. Cleve has recorded the presence of a 

 specimen of lUddulphia aiirita within another, — the internal one 

 being considerably smaller, and having no spines (Bihang till K. 

 Svenska Vet. Akad. Handlinf/ar, Bd. i., No. 13). Besides these 

 modes of reproduction, nothing was certainly known of any mode 

 of multiplication until the observations I was enabled to make in 

 spring and summer on board the ' Garland.' The minute botanical 

 details and considerations have been discussed at length in a paper 

 published, by permission of the Board, in the Proceedings of the 

 Royal Societg of Edinhurgh (vol. xxi. p. 207, plates i.-iii.). 



It was discovered in Biddulphia mobiliensis that the cell-contents 

 contract and round themselves oS, this more or less spherical body 

 proceeding to secrete a membrane of approximately the same shape 

 as the parent, but without the characteristic external spines, &c. 

 This body is apparently of the same character as that recorded by 

 Pi'of. Cleve in B. miiita (of which I was ignorant at the time). 

 During summer, I found these internal bodies free in the water in 

 large numbers, and in a state of active division, none of them 

 having developed the characteristic external spines of the parent 

 form. The parent form was almost wholly absent. In December, 

 again, the parent form (with spines) was abundant, but none of the 

 summer forms. Presumably these had grown into mature Biddul- 

 phia muhiliensis, either directly, as is most likely, or after some other 

 development. The production of this internal form, so far as I 

 have seen, in Biddulphia, is always a case of rejuvenescence of the 

 cell, the whole of the cell-contents of the parent being used up in 

 the formation of one new cell, which, when free, increases its kind 

 by division. I have observed similar rounding off in the cell- 

 contents of Dittjlum Biightwellii, and very probably it goes through 

 a similar life-history. 



In Ooscinodiscus concinnus, I have also found a young form 

 within the parent, this time exactly resembling it in all essential 

 characters. In this genus also, therefore, we have the production 

 of a new individual by the rejuvenescence of the cell. That this 

 often happens in spring I had abundant evidence ; but, what is 

 more interesting, I have found in this species two diatoms within 

 the parent, showing that the protoplasm must have divided to form 

 them. In Loch Fyne I found other specimens of C. concinnus, of 

 which the protoplasm had divided into four, eight, and sixteen 

 rounded-off portions (produced, to judge by their positions, &c., by 

 successive divisions into two) ; and free in the water at the same 

 time, packets of eight and of sixteen young Cuscinodiscus concinnus, 

 of the same size as the rounded portions of protoplasm. The con- 

 clusion was inevitable that the packets of eight and of sixteen were 

 produced from the rounded-off portions, having regard to the 

 occurrence of single forms and pairs as described above. The 



