KOPOID AND SWEZY: UNARMORED DINOPLAGELLATA 29 



the pigment may be more or less broken up and scattered tlirough the cytoplasm 

 f]il. 11, figs. 121, 125; pi. 8, fig. 84). Its wandering movements are also char- 

 acteristic of cytoplasmic pigment. 



This connection between the jjigment of the melanosome and the pigment 

 scattered through the cytoplasm receives some confirmation in observations 

 made by Faure-Fremiet (1914) on Erythropsis pavillardi. The cytoplasm of 

 the form he was working with was of diifuse rose color. Upon the addition of 

 dilute acetone or weak acetic acid to the organism the color was concentrated 

 in clusters of small red glol)ules. The action of acetic acid on the melanosome 

 caused its breaking u]) into small red and brown granules, belonging respec- 

 tively to the outer dark pigment and to the core. The red granules presented 

 a similar aj)pearance to that of tlie "erythrosomes" or precipitated color of the 

 e}i:oj)lasm. 



Faure-Fremiet 's observations also indicate that the pigment composing the 

 core of the melanosome is similar to that found in the cytoplasm. Thus far 

 we have found no evidence in our own work to show that it possesses the mobility 

 characteristic of the outer part of the melanosome. At the time of cytolysis 

 it separates from the melanin as a spheroidal mass which may break up into 

 two parts (fig. TT, 7). Under the action of the sea water it slowly dissolves, 

 as does the remainder of the ocellus. 



The behavior of this organelle at the time of division is totally unknown. 

 In Protopsis ochrea (fig. LL, 4) Wright (1907) has figured a division stage in 

 which the separation of the two zooids has not yet taken place. The ocellus 

 as well as the girdles are fully formed in both daughter organisms. One large 

 body is present in the cytoplasm, l)ut whether this is the still undivided nucleus 

 he does not state. It is probaljly a vacuole, as this stage of formation of the 

 anatomical features of the l)ody would indicate the completion of division of 

 the nucleus. No hint is given as to the mode of formation of the ocelli. Division 

 stages of the ocellate dinoflagellates have been entirely absent from our own 

 material unfortunately, hence we can throw no light upon this subject. 



The fimction of this remarkable addition to the structural characteristics 

 of these flagellates is pi-obleniatical. Tlie reactions of these organisms to light 

 and other stimuli are almost entirely unknown, due to the rarity of the organ- 

 isms and to the difficulties of preserving them under conditions suitable for 

 experimental o1)servation. This organelle is apparently structurally adapted 

 for and may be functionally efficient as a light-perceiving organ, though whether 

 actualh' so remains to l)e determined by experiment. 



Nematocysts. — Of all the organelles of the dinoflagellates two stand out 

 distinctly as metazoan in their type of structure. These are the nematocysts 

 and the ocelli. The latter structures are peculiar to the dinoflagellates alone, 

 while the former have a wider distril)ution in the Protozoa, being found 

 throughout the Cnidosporidia and in at least one species in the Ciliata, Fron- 

 tonia leucas. In the Dinoflagellata they are found in two genera, Polykrikos 

 a.nd Nematodiniiou (Pouchefid (ininitd Dogiel). 



