294 ScioUlfic Proceedings^ Royal Dublin Society. 



met superimposed and twinned with others exhibiting undeniable 

 plagioolastio characteristics. At the same time, many orthoclastic 

 peculiarities are, as will be seen presently, noticeable in the colum- 

 nar crystals. 



Little or no cleavage is apparent in these tabular crystals. 

 Extinction does not occur parallel to the edges of the dominant 

 faces — apparently macropinacoids. Want of properly furnished 

 instruments forbids the author entering at present further into 

 the anomalies presented by these plagioclastic feldspars. They 

 are quite insoluble in boiling hydrochloric acid — or after some days' 

 cold digestion in the same. 



The second variety, if it may be considered as such, may be 

 described as columnar, with nearly equally developed faces seem- 

 ingly at right angles to each other. Four pyramidal faces termi- 

 nate one or both ends. This pyramidal termination is occasionally 

 truncated, giving a fifth terminating face at an angle other than 

 90° with the zone axis of the prismatic faces. Marked rectangular 

 cleavage on one set of these faces ; direction of cleavage apparently 

 parallel to pyramidal planes : generally a single medial twinning 

 line on the other set of prism faces. Extinction for the first set of 

 faces (those showing cleavage), parallel to the longitudinal axis of 

 the crystal. For the second set, symmetrically on either side of 

 twinning line; seldom more than 15° on either side. A specimen 

 wanting the twinning line extinguished on these faces at 16° with 

 the longitudinal axis ; parallel to it on the adjoining faces of the 

 same zone. 



Viewed when resting on one of its longitudinal edges, charac- 

 teristic border lines displaying vivid prismatic colours become 

 apparent between crossed nicols. Polysynthetic surface striations 

 might account for this appearance. 



The photograph (PL xii. fig. 1), taken with nicols crossed, con- 

 tains a couple of these crystals removed from the floating pumice. 

 The crystals, adhering together, rest on their edges and display 

 well the border lines just mentioned. (Figure 1a. PI. xii.) 



Flame tests indicate for such crystals a percentage of potassium 

 comparable to that of orthoolase. If they, indeed, are sanadine, 

 there is evidently a peculiarity in twinning. Although good crys- 

 tals are scarce, fragments are numerous. They will be detected 

 by their marked cleavage. 



