26 Florida Agricultural Experiment Station

or narrow range in percentage composition as being representa-
tive for a given vegetable crop in the state.
 High protein percentages in the crops were associated with
 organic soils such as those of the Belle Glade area. High calcium
 and magnesium were associated with calcareous soils or those
 relatively high in exchangeable calcium and magnesium.
 On a dry weight basis, cabbage showed significant differences
 in percentages of protein, calcium, magnesium, potassium, phos-
 phorus and iron between certain areas in the state; beans, dif-
 ferences in protein, calcium and phosphorus; celery, differences
 in calcium, potassium and phosphorus; and tomatoes, differences
 in protein, calcium, potassium and phosphorus.
 Compared with the general average of all areas sampled, cab-
bage samples from the Belle Glade and Hastings areas were
above average in iron, while those from Bradenton and Winter
Garden were below average in this element. Cabbage leaves
averaged higher in all constituents than did the heads.
 Beans from the Belle Glade area were below average in phos-
phorus, those from the Perrine marl at Homestead were above
average. Those from the Palm Beach area were below average
in calcium.
 Celery from the Belle Glade area was above average in potas-
sium while that from the Sarasota area was above average in
phosphorus but below average in calcium.
 In calcium, tomatoes from the Homestead area were above
average, those from Ft. Pierce and Ft. Myers areas below aver-
age. Samples from the Ft. Pierce area were above average in
phosphorus and those from Collier County were below average
in this element.
 Collard leaves averaged 86 percent more calcium and 47 per-
cent more iron than the petioles. Collard leaves were higher in
percentage protein, calcium, magnesium and iron than the outer
cabbage leaves.
 There was little correlation between fertilization or soil an-
alysis and plant composition for a given area of similar soils.
The factors that characterized soil types appeared to be organic
matter content and pH of the soil. Other factors of soil environ-
ment and moderate differences in fertilization were of secondary
importance.
 The average mineral composition of Florida vegetables was
as often above as below that reported for other sources. Ex-
tensive data of the type reported herein for Florida must be ob-