The Working of the Madison Lakes. 123 



season advanced, the larger Anabaeiia became relatively a trifle more abun- 

 dant, but for the most part remained sterile. 



Examinations were not made for several weeks in August and September, 

 but on the 26th of September, when another sample was taken for the use 

 of a laboratory class, the bloom consisted exclusively of the smaller Ana- 

 haena (Fig. 5), which I have called A. mendotae,^ a species that continued 

 to ai)pear on the shore at intervals until well into November. 



M. Bornet, whp was obliging enough to examine specimens of this alga 

 for me, writes as follows: " The Anabaena you have sent me belongs to 

 the sub-genus Doliclwspermum Thwaites, in the sense in which it is 

 understood by Wittroch (Wittroch and Nordstedt: Algae exsiccatee. No. 496). 

 It appears to me new, and differs from all of the species which constitute 

 water-bloom in the tenuity of its threads, in the length of its articles and 

 the slenderness of its spores. It is only compai-able to LimmocJilicle flos 

 aquae in the diraensions and configuration of the dilTerent vegetative and 

 reproductive organs. That you may compare it, I send a tracing from 

 analyses of authentic specimens of Nostoc flos aquae Lyngbya [Fig. 1], 

 Nostoc flos aquae Agardh [Fig. 2], and Anabaena circiiialis Rabenhorst 

 [Fig. 3], which is the prototype of the species. Admitting that these three 

 plants are but forms of a single species, as may be inferred from the ob- 

 servations of Wittroch on Anabaena {Doliclwspermum) hassallii, your alga 

 still differs from all of them, and is, especially, very distinct from the Ana- 

 baena circiiialis of Rabenhorst." 



As a general tlaing, i^rotracted heat precedes the appearance of the 

 ■water-bloom, which is most frequent in sluggish or stagnant bodies of 

 water, especially such as are fed by the outlets of bogs, hence the phenom- 

 enon is most frequently noticed in mid-summer or later: yet the abundance 

 of the scum in June 18S2, shows that a very short period of warmth may 

 suffice for its formation, as the early part of the spring was cold, and there 

 had l)een but few liot days when it appeared. 



Unlike the majoritj^ of the algae implicated, Limnochlide flos aquae 

 thrives exceptionally well in cold water, and has even been found luxur- 

 iating in the greatest abundance in ice.' Anabaena mendotae, which first 

 became really abundant after the disappearance of Clathrocystis coelos- 

 phaerinm and the larger Auabaeiici, and lasted until not only the air but 

 the lake itself had become quite cold, appears to resemble it somewhat in 

 this respect. 



During the summers of 1883 and 1834, I was absent from the city, but I 



^Anabaena iDolichospermum') mendotae, n. sp — Threads circinate. Vegetative cells 

 slender, usually elongated, especially towards the ends of the filaments, 3-4 X 4.5-12//. 

 Heterocysts ellipsoidal or barrel-shaped, 4.5 X (J-~-'3 M . Spores remote from the heterocysts 

 slender, more or less curved, 4.5-7.5 X 20-40 ju , bluish-green, Uke the vegatative cells (Fig. 5) 

 Forming a copious water-bloom on Lake Mendota, at Madison, Wis., especially abundant 

 in the faU. 



-Magnus: I. c. 



