294 Varietäten, Descendenz, Hybriden. 



of O. lamarckimia grown in the New York Botanical Garden in 

 1904 was sown in the garden in August 1904, furnishing a total of 

 499 plantlets. Among these the mutants were: O. nanella 3, O. lata 

 2, O. oblonga 3, O. albida 6, Unidentiiied 5. 



Culture D. 1. 9. Purely fertilized seeds of 0. laniarckiana har- 

 vested in the New York Botanical Garden in August 1904, were 

 sown in November of the same 3''ear. 604 plantlets were produced. 

 The mutants were: O. nanella 1, O. lata 2, 0. scintillans 2, O. albida 

 1, 0. oblonga 2, Not indentitied 11. 



Culture from Bicknell's seed. Seeds coUected from the herba- 

 rium specimens of 0. laniarckiana collected hy Mr. Bicknell at 

 Nantucket City were planted. Twenty-four of the most diverse 

 plantlets which could be selected from the lot of seedlings were 

 transplanted and proved to consist of 6 individuals apparently be- 

 longing to 0. albida, although 4 of them died early and the other 

 two did not get be3^ond the rosette stage in 1905, and 18 typical 

 O. lamarckiana. 



Culture from Vilmorin's seed. 203 plants grown from com- 

 mercial seed produced 1 O. nanella, 1 0. lata and 1 0. albida. 



Culture from English seeds. Plants grown from seeds of the 

 English "O. biennis''' collected at Bidston Junction by Mr. C. T. 

 Green proved to be indubiable O. laniarckiana. Among a prepon- 

 derating number of the parent species 2 individuals of O. lata and 

 4 of O. rubrinervis were noted. The 2 O. lata plants were more 

 robust than SLny hitherto studied and were the first to yield pollen 

 sufficient for fertilization. 



Culture from self-fertilized O. lata. From the two individuals 

 of O lata mentioned above it was possible to secure a few capsules 

 producing a small quantity of seed from fertilization with O. lata 

 pollen. From this seed were rai.sed: O. lamarckiana 80, 0. lata 10, 

 O. oblonga 3, 0. albida 1. 



From these experiments the authors conclude that the coefticient 

 of mutability has not been increased in the American cultures. 

 One new form which was not brought to maturity was discovered. 



O. lata, O. oblonga, 0. albida, O. scintillans and O. brevistylis 

 were grown from pedigreed seed and detailed taxonomic descrip- 

 tions and excellent habit and detail figures are given. 



3) Biometrie Investigations of Variation in O. lamarckiana and 

 its mutants. 



Shull, in continuing his Statistical studies, has emploj^ed 6 

 series of material, all cultivated at the Station for Experimental 

 Evolution but grown from seed of very different origin. The selection 

 of series of material of widely different origin affords a very critical 

 test of the validity of the specific characters in question. 



As in his earlier research, he found that the vegetative charac- 

 ters were very subject to the influence of environmental factors. The 

 bud, however, seems a satisfactory subject for biometric investiga- 

 tions. It is divided into three very distinct and easily measured 

 regions, the ovary, hypanthium and the cone. 



The biometric data are interesting from two points of view: a) 

 as and aid to the evaluation of specific differences, and b) in their 

 bearing upon general biological problems. 



For the seriations of original measurements and the constants — 

 means, Standard deviations, coefficients of variability, and coefficients 

 of correlation — the original paper must be consulted. A table 

 showing the percentage deviation of the means and variabilities of 



