NOTES AND ABSTRACTS. 



633 



these researches, and a large number of species were crossed both 

 ways. A certain number of second-generation hybrids were obtained. 



Not only the original species, but the hybrids are described in 

 considerable detail. The anatomical structure was specially examined 

 and is often figured. 



The author finds that the hybrids are more usually homogeneous, 

 but in eight cases are heterogeneous; well-marked differences occur 

 in three sets of reciprocal crosses. There are all sorts of transitions, 

 both between brother hybrids and reciprocal crosses, some resembling 

 each other closely and others showing great differences. The differ- 

 ences often resemble those that characterize Jordan's species. 



In crosses between those species which have five and those which 

 have three sepals the majority have five sepals, a very few have three 

 sepals, but others have four sepals, and occasionally forms with six 

 and seven sepals occur. In crosses of C. ladaniferus (ten loculi) with 

 others (five loculi) he found in one set of sixty-five capsules, fifty 

 with five, fourteen with six, and one with seven loculi. The species 

 used differ considerably in the character of the hairs, which vary from 

 stellate to simple. These were inherited differently. Intermediate 

 forms of hair occurred in thirteen out of eighteen cases, sometimes 

 with, sometimes without the characteristic form of one of the parents ; 

 or in a very few cases only one type of hair was transmitted, or 

 sometimes both kinds occurred together in the hybrid. 



In a few cases a single branch of a hybrid resembled the father, 

 whilst another branch of the same plant took after the mother. 



Occasionally one character appears identical with that of one of 

 the parents, but examples of heredity from both parents are more 

 frequent and more distinct. 



The sterility of the pollen was in most hybrids a very marked 

 characteristic. In one case only the pollen seemed to be of a normal 

 character. In many cases all the pollen-grains were empty, and 

 in most hybrids at least 70 per cent, were so. 



The spots on the petals of C. ladaniferus var. maculatus were 

 sometimes strongly marked in the descendant, but in one case were 

 not transmitted at all, and in other cases differed greatly in size in 

 different individuals. 



The characters transmitted by a species vary with the mate 

 selected, according to the affinity of the two species. The predominance 

 of the mother was clear in seven and of the father in two cases. In 

 vine hybrids the predominance of the pollen parent is the general rule. 



These instances are sufficient to show the importance of this 

 paper. — G. F. S. E. 



Coconut Bud-rot, The History and Cause of the. By John 



E. Johnstone {U.S.A. Dep. Agr., Bur. PI Ind., Bull. 228; 1912).— 

 The author of this paper claims to have shown that the organism 

 giving rise to Bud-rot of Coconut is identical with Bacillus coli of 

 animal origin. 



