demi-stauros. Frustule constricted at the centre, ends truncate, connecting zone smooth. Length 55 to 74 μ, breadth 17 μ 6. Foveaux Strait. Pl. XIV., Fig. 15.

This is a remarkable form; in the s. v. it appears to be allied to N. Richardsoniana,* from which it differs by the terminal nodules being placed a little below the apices, and by the blank space on one side of the central nodule. [In this respect resembling Pfitzer's genus Anomæoneis.-F. K.]

12. N. rhombus (P. Petit). Valves conspicuously rhomboid, apices acuminate and slightly produced, median line straight, nodules very small, striæ decussating and very close. Length 36μ 6, breadth 24 μ 2; about 35 striæ in 25 μ.

Foveaux Strait. Pl. XIV., Fig. 12.

This little species is remarkable for its rhomboid form and crossed striæ, which under a high power form lozenges arranged in quincunxes. It is very closely allied to the species described and figured by Lewis † as N. placentula, Ehr. [This is not N. placentula, but N. rostellum, W. S., Sy. Brit. Diat.,' vol. ii. p. 93 = N. apiculata, Greg., 'Quart. Jour. Mic. Sci.,' vol. iv. p. 4, pl. i. fig. 13. I have seen authentic specimens of this form; it is not uncommon in fresh-water gatherings from Morven, Aberdeenshire, communicated by my friend, the Rev. G. Davidson of Logie Coldstone.-F. K.]

STAURONEIS, Ehr.

13. S. robusta (P. Petit). Valves elliptical, suddenly acuminate towards the ends; median line straight; central nodule large, terminal nodules small and placed a little below the apices; stauros dilated towards the margin, but which it does not reach; striæ interrupted by marginal blank spaces; margin annular, striated; frustule strongly contracted at the centre, and largely rounded at the summits; connecting zone smooth. Length 90 to 101μ, breadth 28μ. Striæ, 18 in 25 μ.

Foveaux Strait. Pl. XV., Fig. 16.

This species is distinguished from all others by its suddenly acuminated apices, the position of the terminal nodules, its annular margin, the interruption of the striæ, the contraction of the frustules at the centre, and their rounded apices.

[This form undoubtedly belongs to the group of which Navicula aspera (S. pulchella, W. S.) is the type. The forms have no true stauros, but only a median blank space, which is sometimes dilated, and sometimes circular.-F. K.]

AMPHIPRORA, Ehr.

14. A. rugosa (P. Petit). Valves in f. v. narrow, contracted at the centre; apices terminating in curved points; striæ represented by * O'Meara, 'Irish Diat.,' p. 339, pl. 32, fig. xxxiii.

U. S. Diatoms,' part iii. p. 8, pl. ii. fig. 7.

puncta irregularly disposed, and giving the valves a rugose appearance. Length 132 μ.

Campbell Island (rare). Pl. XV., Fig. 17.

[This form belongs to my genus Perrya.-F. K.]

SURIRELLA, Ehr.

15. S. Filholii (P. Petit). Valves elliptical; margin contract; extremities largely rounded; costa dilated towards the margin, not reaching to the centre; each of these has two or three firmly punctate striæ of the same length as itself. Midway between the summits and the centre are two short parallel punctate lines (vers le centre se trouvent deux lignes punctuées et atrophiées). Length 108 to 126μ, breadth 42 to 54 μ.

Foveaux Strait. Pl. XV., Fig. 18.

This beautiful species (dedicated to M. le Dr. H. Filhol) is allied to S. arabica, Grun., in Schmidt's' Atlas,' pl. xx. fig. 5; but it does not possess the uncontracted second margin of that species; its costæ are also more distant. Its place is between S. fastuosa and S. arabica.

TRACHYSPHENIA,* P. Petit.

This genus by its cuneiform valves establishes its connection between the Fragilarias and Meridion. The frustules, as seen in f. v., are quadrangular. The cuneiform valves are covered with puncta disposed in vertical and horizontal lines at right angles to each other. The cost of Plagiogramma do not exist here.

16. T. australis, P. Petit. Characters same as genus. Length 35 to 52 μ, breadth 8 to 11 μ.

Lyell's Bay; Campbell Island. Pl. XV., Fig. 19.

[The valves can scarcely be called cuneiform; they are ellipticlanceolate, and the ends are usually symmetrical. O'Meara refers a similar form to the genus Terebraria, of Greville (T. Kerguelensis, Linn. Soc. Jour.,' vol. xx. p. 56, pl. i. fig. 4). It is perhaps identical with Fragilaria pinnatula, Ehr., Mik.,' pl. xxxv. A, xxii. fig. 8. T. australis differs from T. Kerguelensis by its larger granules and the blank spaces at the apices.—F. K.]

GRAMMATOPHORA.

17. G. longissima (P. Petit). Frustular view narrow; undulations of the villi nearly reaching the centre; number of undulations 25. Length 114 μ, breadth 8 μ 8.

Lyell's Bay (rare). Pl. XV., Fig. 21.

This curious species is distinguished from G. serpentina by the narrowness of the valves, and the extremities of the undulation

*Tpaxús, rough; and Zohv, a wedge.

being straight as in G. marina. I have been unable to obtain a s. v. of the diatom. I, however, believe them to be arcuate.

RHABDONEMA.

18. R. hamuliferum (F. Kitton). Valves lanceolate, undulate, with transverse punctate striæ, ends rounded smooth, diaphragms* striate their whole length, having in their centre a circular ring, and towards the extremities two others much smaller and elliptical. Seen in f. v. the frustules appear composed of a variable number of diaphragms (annuli) placed between the two valves, and separated the one from the other by lines in the form of hooks; each of the diaphragms is striated on the edge. Length 33 to 78 μ, breadth 11 to 15 μ.

μα

On corallines, Lyell's Bay. Pl. XV., Fig. 22.

This form is distinguishable from all other species of Rhabdonema by the undulations of the margin of the valves, and by the lines separating the diaphragms being bent in the form of hooks. Mr. Kitton recognized it as new, and named it hamuliferum.

AULISCUS, Ehr.

19. A. stelliger (P. Petit). Valves circular, with four processes, two of which are very small. The valve is divided into three concentric parts, the innermost has five stellate rays, the intermediate portion has the rays irregularly disposed, the external circle has a row of puncta on the inner margin corresponding with the rays of the intermediate part; in this outer ring are situated the four processes, between which are irregular rays. Diameter 35 μ 2.

Campbell Island (rare). Pl. XV., Fig. 25.

This species is distinguished from all other Aulisci, recent or fossil, by the peculiarity of the disk and by the presence of the interrupted rays.

The following genera and species of other authors are figured by M. Petit:

1. Cocconeis Wrightii (O'Meara), 'Q. M. J.,' vol. vii. n. ser. p. 246, pl. vii. fig. 6.

Lyell's Bay. Pl. XIV., Fig. 3.

The figure here given represents a very small form, not exceeding 13 μ to 26μ 4, but the difference from the type form is too little to allow of its being considered a variety; by altering the focus of the objective a dilated cruciform nodule may be detected.

The so-called diaphragms as figured by M. Petit are valves; the diaphragms (annuli) in this genus are only striate on the edges. The "hooks" represent the septa as seen in s. v.

2. C. Crux (Ehr.), 'Monatsberichten,' p. 265.
Lyell's Bay. Pl. XIV., Fig. 4.

3. Campyloneis Grevillii (Wm. Sm.).

Since the creation of the genus by Herr Grunow, he has discovered that C. Grevillii consists of two valves; the inferior is composed of two superimposed plates of silica, of which no drawing has previously been given. Fig. a represents a complete frustule; Fig. d, the superior valve; b and c, the two plates of silica composing the inferior valve.

4. Navicula decussata? Ehr. Bericht. Berl. Akad.' p. 364. Length 30 to 39 μ, breadth 7 μ 7.

Lyell's Bay. Pl. XIV., Fig. 11.

This species with crossed striæ has not been previously figured; I therefore doubtfully refer the present form to this species. Nevertheless all the characters correspond with those given by Ehrenberg.

5. N. Hennedyi (Wm. Sm.), var. B constricta.

Foveaux Strait. Pl. XIV., Fig. 13.

This form has probably not been figured; it is chiefly remarkable on account of its occurrence with a species essentially elliptical, clearly showing that the contraction of the margins is of no value as a specific distinction.

6. N. Smithii (Breb.), var. 8 minor (P. Petit), n. var.

This variety is distinguished by its broadly rounded ends, and its finely punctate striæ. Length 33 μ, breadth 19 μ 8, costæ 15 to 16 in 15 μ.

Lyell's Bay. Pl. XIV., Fig. 14.

7. Grammatophora marina (Lyng., n. var. (P. Petit).

This variety differs from the type, by the lines of the diaphragms being more bent.

Foveaux Strait. Pl. XV., Fig. 20.

New Diatoms from Honduras. By Herr GRUNOW.

'Monthly Microscopical Journal,' October 1, 1877.

ERRATA.-PL. CXCIII., for 11 a b c d, read 10 a b c d; for 10 a b c, read 11 a b c.

III. Further Remarks on a "Simple Device" for the Illumination of Balsam-mounted Objects for Examination with Immersion Objectives whose Balsam Angle is 90° or upwards. By J. J. WOODWARD, Surgeon and Brevet Lieut.-Col. U.S. Army.

(Read before the ROYAL MICROSCOPICAL SOCIETY, October 9, 1878.) SINCE the publication of the original paper in the 'Monthly Microscopical Journal' for August, 1877 (p. 61), I have been in the habit of using a similar prism to that there described, without any diaphragm, in the ordinary lamplight illumination of balsammounted tests for study with objectives of the kind referred to. The apparatus is so simple, so easily manipulated, and so effective, that a brief description may be of interest.

с

a

A right-angled prism of crown glass (a), the long side of which is three-quarters of an inch long by half an inch wide, has its right angle truncated as in my "simple device," and cemented to a small base-piece of brass (b) which is supported in a stiff steel rod (c) three or four inches long, and about the thickness of a knitting needle. The whole apparatus ought not to cost more than three or four shillings.

To use it, I slip the steel rod into the holder of the rod of the dark well usually employed with the Lieberkühn mirrors. This holder fits into the sub-stage of the microscope, and putting a drop of oil of cloves on the upper face of the prism, it is racked up into optical contact with the lower surface of the slide. I then condense the light of a small coal-oil lamp upon the object, through either face of the prism, by a small bull's-eye lens of about three inches focal length. One half of this condensing lens may be covered with black paper, in which case it can readily be so placed that no ray forming a smaller angle with the optical axis than 45° can enter the objective. Excellent results can be obtained, however, without this precaution, if the intention be merely to resolve a given test, without caring to know precisely the angle at which this is done; in this case it is only necessary to place the lamp in such a position that a line drawn from its flame to the face of the objective shall be perpendicular to either short side of the prism, and condense the light in that line. By looking through the other short side of the