906 



SCIENCE 



[N. S. Vol. XXIX. No. 753 



varieties 1 part in 80,000 causes the closing of the 

 open flowers upon twelve hours' exposure. This 

 injury takes place directly on the bud or flower 

 exposed and not indirectly through absorption by 

 the roots. No chemical test is delicate enough to 

 detect the least trace of illuminating gas in the 

 air. Ethylene is even more fatal to the flowers 

 of the carnation. Three days' exposure to 1 part 

 in 1,000,000 prevents the opening of buds just 

 showing the petals. Twelve hours' exposure to 1 

 part in 2,000,000 causes the closing of flowers 

 already open. There is much evidence that indi- 

 cates that the toxic limit of illuminating gas 

 upon these flowers is determined by the ethylene 

 it contains. 



Types of Cactus Genera, 1753-1904: J- N. Rose, 



Smithsonian Institution, Washington, D. C. 



Linnaeus in 1753 referred all the cacti known 

 at that time to one genus, which he called Cactus. 

 Otto Kuntze in 1904, one hundred and fifty years 

 afterwards, recognized but three genera; two of 

 these, Pterocactus and Pereskia, are small or con- 

 tain less than a dozen species, and hence the great 

 mass of cacti are to-day to be found in the genus 

 Cactus as understood by Linnaeus. This might 

 indicate that Linnseus's conclusions were good, 

 but, as we all know. Otto Kuntze was a poor 

 botanist, although he was a great bibliographer 

 and doubtless did more than any one in modern 

 time to stimulate botanical bibliography. Since 

 Linnaeus's time 58 genera have been proposed. 



What are the types of these genera? Natural- 

 ists are now all agreed that the type of a genus 

 must be one of the original species in it. At the 

 present time, however, we have two genera of 

 cacti which do not contain any of the original 

 species and of course are used in an entirely dif- 

 ferent sense from that which was first intended. 

 These genera are Pilocereus and Epiphyllum. 



Botanists are now pretty well agreed that we 

 ought not to use a homonym of an older genus, 

 and yet we have at the present time in cacti two 

 names which come under this class, viz., (1) 

 Earriota DC, 18 — , while there is Harriota 

 Adans, 1762, also a cactus; (2) Mamillaria Haw, 

 1812, while there is Mamillaria Stackh, 1809, a 

 genus of algje. 



Botanists are now pretty well agreed through- 

 out the world that our nomenclature should begin 

 with 1753 and that older names shall not displace 

 Linnaean and post-Linnaean names, but at the 

 present time we have in cacti one pre-Linnaean 

 name which was not taken up until 1827, viz., 

 Melocactus. 



In the paper as presented there follows an 

 alphabetical list of the genera of cacti, showing 

 the date of publication of the genus, the number 

 of original species and the names of the type 

 species. 

 Some Variations and Hybrids of (Enothera: E. 



R. Gates, University of Chicago, Chicago, 111. 



The extreme variant of 0. rubrinervis which 

 appeared' in my cultures is found to breed true 

 to its peculiarity. The presence of red on the 

 hypanthium and in excess on the sepals is cor- 

 related with its development in excess on the 

 under surface of the rosette leaves. The same 

 correlation exists in one of the types from 0. 

 nanella X 0. biennis. The 0. biennis in these 

 crosses was from the type growing wild around 

 the New York Botanical Garden. Seven plants 

 germinated from this cross and five of them 

 reached maturity. These were of two sorts, four 

 of which belonged to one type. In this type the 

 rosette leaves were long, rather narrow and 

 pointed, like 0. rubrinervis. The petioles were 

 red above but more conspicuously so on the under 

 surface. The buds were large like 0. Lamarckiana 

 and, like the extreme variant of 0. rubrinervis 

 mentioned, the hypanthium and sepals were red 

 throughout. The excessive development of red 

 pigment from a cross between two types, neither 

 of which shows much red, is an unexpected result 

 several explanations of which are possible. In the 

 second type the buds were small and greenish, 

 showing the 0. biennis characters. 



0. Lamarckiana X 0. biennis. Nine plants ma- 

 tured, all of one type. The rosettes were very 

 much like 0. nanella X 0. biennis, type I, but 

 larger, and the petioles were bright red above but 

 without red on the under surface. The buds 

 showed the 0. biennis eharacters, being small, 

 with little red on the sepals and with a short 

 style. The petals varied in size from that of 

 0. biennis to intermediate between 0. biennis and 

 0. Lamarckiana. 



These crosses were made by Dr. D. T. Mac- 

 Dougal, who presented the seeds to the writer for 

 further cultures. 



An interesting " combination type " appeared in 

 a culture of English evening primroses which con- 

 tain some of the mutants of 0. Lamarckiana and 

 in addition some new types. The type in question 

 showed the eharacters of 0. Lamarckiana in its 

 rosette and stem leaves (absence of red) but the 

 buds were typical 0. rubrinervis. This eombina- 



'■ See Science, 27 : 209, 1908. 



