1917] FIELD CROPS. 333 



In summarizing the results the author states that the grasses studied have 

 theii* main blossoming stage between five and nine o'clock in the morning, with 

 the exception of LoUum perenne and L. italicum which blossom a little later, 

 or between ten and twelve o'clock. The blossoming of an individual floret gen- 

 erally lasts from one to two hours, seldom less than one hour but often longer 

 than two hours. No marked difference in this respect among the different 

 species was established. 



Blossoming was observed to begin in the upper third of an inflorescence, 

 and in the branches of panicles the flower at the point was the first one to 

 open. With the exception of Poa pratensis, the blossoming began with the low- 

 est floret and proceeded upward. It was found that on the average the earlier 

 blossoms produced the heavier fruits. Weather conditions prevailing on a 

 certain day, or even the day preceding, were found to exert a marked influence 

 on the beginning of blossoming and the number of blossoms opening. A re- 

 tarding effect of the weather gave rise to a larger number of opening blossoms 

 on the day following. 



Experiments in the artificial control of temperature, light, and moisture 

 showed that temperature was of much greater importance in its effect on blos- 

 soming than the other two conditions. With insufficient temperature blossom- 

 ing did not take place, but with the proper temperature conditions it was only 

 slightly retarded by the absence of light. High atmospheric humidity did not 

 interfere with blossoming, although it held back the opening of the pollen sacks. 



Single blossoms of different gi-asses inclosed in cotton batting did not pro- 

 duce seed-containing fruits. Similarly inclosed single inflorescences, in the 

 case of only a few grasses, yielded negligible quantities of gerniinable seed, and 

 two inflorescences from an individual plant produced sexually and inclosed 

 together yielded some fruits which did not germinate and others, somewhat 

 greater in number than those secured from the singly inclosed inflorescences, 

 which were capable of growth. Entire sexually-produced plants when in- 

 closed did not, in many instances, yield a single germinable seed, in others 

 only a small number, and in a few cases a larger number. L. perenne, L. itali- 

 cum, and P. elatius exhibited the greater tendency toward self-fertilization. 

 The great differences in the behavior of plants of the same species under this 

 treatment are ascribed to individual differences. 



Divisions of sexually produced plants inclosed together resulted in a num- 

 ber of cases in the development of a satisfactory number of germinable seeds. 

 The progressive results secured in this series of experiments are considered due 

 to the greater chances of pollination in any particular test as compared with 

 the preceding one. 



Experiments with divisions of an individual plant of A. elatius showed that 

 complete self-sterility occurs in some grasses. For two years in succession 

 297 such divisions planted 40 cm. apart each way did not produce any ger- 

 minable seed. 



Some common grasses and how to know them, R. S. Macdougaxl (Trans. 

 Highland and Agr. Soc. Scot., 5. ser., 28 {1916), pp. 56-91, figs. 28). — ^This paper 

 describes the root, stem, leaf, flower, fruit, and seed of a grass plant, defines 

 the terms used for descriptive purposes, gives a detailed description of each 

 of the more important grasses together with keys in tabular form for their 

 identification, and points out how the seeds of different species may be 

 recognized. 



Eleven references to literature on grasses are cited. 



A trial of grass mixtures, F. W. Hilgendorf {Jour. Canterbury Agr. and 

 Past. Assoc, 3. ser., 4 {1916), pp. 28-34, figs. 5). — A comparison was made on 

 land of medium fertility of grass mixtures consisting principally of Italian 



