ZOOLOGY AND BOTANY, MICROSCOPY^ ETC. 511 



the auxospores are always in contact with one another in pairs. With 

 the " box-hypothesis," the theory that the individual is continually 

 decreasing in size must also be abandoned. The young cell-wall of 

 the newly formed cell is very soft and extensible, and readily attains 

 the size of that of the mother-cell. 



The auxospore germinates by a simple process of cell-division, 

 commencing with a division wall at right angles to the axis of 

 growth, 



Friishdia saxonica j)resents another favourable example ; but in it 

 also Hallier determined the incorrectness of the " box-hypothesis." 

 The process of cell-division differs, however, in some important 

 points from that in Melosira. There is no splitting of the cell-wall 

 of the mother-cell ; the division is, on the contrary, preceded by an 

 extension of the cell-wall in the direction of the transverse diameter 

 of the secondary side, and a cell-wall is then gradually formed 

 dividing the individual into two halves along its longest diameter. 



With respect to the movements of FrustiUia, Navicula, and other 

 similar forms, the author makes the following observations. The 

 movement usually takes place on the primary side, one end being 

 somewhat depressed, the other end being somewhat elevated and 

 moving backwards and forwards, the individual gradually advancing. 

 Two incorrect assumptions have, however, been made in order to 

 explain this movement ; viz. that the motion is always a gliding one, 

 and that it is caused by a crevice on the primary side (which was 

 even alleged to have been seen in Navicula and Frustulia), through 

 which the protoplasm rotating in the interior of the cell is in com- 

 munication with the outer air. 



Neither Navicula nor Frustulia shows, with the highest magni- 

 fying, the least indication of a crevice ; nor is the rotation of the 

 protoplasm of such a nature as in the least to account for a movement 

 backwards and forwards of the frustule, being quite of the same nature 

 as that which so commonly occurs in living vegetable cells, and 

 affording no explanation of the power of movement of the cell. 

 The following explanation of these movements is offered by the 

 author. 



When the movement is carefully watched for a considerable period 

 and in a number of specimens, it is seen to be subject to a great 

 variety of modifications. If the movement were of a gliding nature, 

 and due to protoplasm exuded through a crevice, the primary side 

 must always face the observer, because in this position only could the 

 rotating protoplasm be in contact with the glass plate. So far, how- 

 ever, is this from being the case, that the diatom is frequently seen 

 to be progressing as quickly on its secondary side, and very often the 

 position is constantly shifting. It is sometimes fixed to the glass 

 plate by one of its viscid ends, swaying backwards and forwards, a 

 phenomenon quite inexplicable on the old hypothesis. In some 

 species, and especially in Cymbella, violent convulsive movements are 

 frequently seen, and even the swaying motion is by no means always 

 very regular. All attempts hitherto to explain the phenomenon have 

 proceeded on the supposition that the " lorica " or cell-wall of the 



2 M 2 



