234 N. E. GREEN ON THE SURFACE MARKINGS OF DIATOMS. 



greater than the projection of one bead from another. (See fig. 6 ) 

 If this observation be correct, an important fact is developed in 

 connection with the form of these beads or pellets in P. Angula- 

 tum, viz., that the curved top of the pellet surmounts some other 

 shape, which may be cylindrical, but in all probability is hexagonal 

 — at least, at the point of contact with the smooth surface of the 

 valve. If so, the old and now somewhat rejected idea of the hexa- 

 gonal markings may be true after all. Let us examine this 

 supposed form by the appearances presented during the opera- 

 tion of focusing. When a valve of P. Angulation is viewed 

 under a y 1 ^, in the usual manner by transmitted light, as the ob- 

 ject glass is gradually brought down upon it, the first set of mark- 

 ings that the eye can recognise as being in focus, are a series of 

 ghost-like dots, very small and evanescent; they come and go with 

 the least touch of the slow movement. A little nearer, and rows 

 of dots of a decisive character are seen ; these are followed by 

 hexagons, which maintain their form for some time, then pass into 

 a second series of firm dotted lines, and" finally the ghost-like points 

 show for a moment, and all is gone. Now r , comparing these appear- 

 ances with the facts revealed by side illumination, we may arrive at 

 this solution : — The first faint dots are produced by the tops of the 

 pellets, the .firm dotted lines by the shades of their sides ; the 

 hexagonal form is due to the shape of the pellet at its contact with 

 the plane surface of the valve, its continuance being accounted for 

 by the repetition of this shape in the pellets on the other side (see 

 fig. 7) ; the second series of decided dotted lines is produced as 

 before by the shade on the side of the projecting portion of the 

 pellets, and the last faint set of dots by the summits of their curved 

 surfaces. 



It should, however, be observed with regard to the amount of cur- 

 vature in the top of the pellet, that although the powers employed, 

 together with the illumination by the lime light, brought out in the 

 most unquestionable manner the beaded surface of Hippocampus, 

 Cocconema, and Cymbella, &c, and even exhibited it on such 

 minute forms as Cluthensis, it required the greatest care, both in 

 handling and in seeing, to produce it in Angulation. 



With regard to side illumination and reflected light versus under 

 illumination and transmitted light, the first is valuable only for sur- 

 face markings, but for these it is invaluable, as the following state- 

 ment will demonstrate : — A valve of Triceratium was under exami- 



