630.7 I£6b no. 675 cop. 8 UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN AGRICULTURE . FLORICULTURAL SALES IN MASS MARKET OUTLETS by R. A. Kelly . BULLETIN 675 UNIVERSITY OF ILLINOIS • AGRICULTURAL EXPERIMENT STATION CONTENTS Page Experimental Procedure 6 SALES: MASS MARKETS AND FLORAL SHOP Retail Sales 9 Method of collecting data 9 Limitations 9 Trend in weekly sales 11 Total sales 12 Average weekly sales 12 Variation in Weekly Sales at Each Outlet 13 Limitations 13 Sales 15 Holidays and special occasions 28 Competing products and displays 31 Changes in display location 32 Addition of various kinds of plants to the display 33 Shrinkage 34 Sales by color and type 38 RETURNS: EXPERIMENTAL AND OTHER MERCHANDISE Limitations 39 Gross retail margins 40 Weekly gross returns per square foot of display space 42 Weekly gross profit per square foot of display space 42 SUMMARY AND CONCLUSIONS . . .43 See nexf page for list of tables and charts. Urbana, Illinois August, 1961 Publications in the bulletin series report the results of investigations made or sponsored by the Experiment Station TABLES AND CHARTS Page Trends in weekly floricultural sales by outlet 1955-56, 1956-57, and 1 957-58 (Fig. 1 ) 10-11 Table 1. Floricultural sales by outlet and year 13 Total number of sales and percent, by outlet and year (Fig. 2) 13 Weekly average sales, by outlet and year (Fig. 3) 14 Table 2. Weekly sales by type of outlet, 1955-56 16 Table 3. Weekly sales in supermarket by kind of plant, 1956-57 17 Table 4. Weekly sales in variety store by kind of plant, 1956-57 18 Table 5. Weekly sales in floral shop by kind of plant, 1956-57 19 Table 6. Weekly sales, all outlets, by kind of plant, 1956-57 20 Table 7. Weekly floricultural sales by outlet, 1956-57 21 Table 8. Weekly sales in supermarket by kind of plant, 1957-58 22 Table 9. Weekly sales in variety store by kind of plant, 1957-58 23 Table 10. Weekly sales in floral shop by kind of plant, 1957-58 24 Table 1 1. Weekly sales, all outlets, by kind of plant, 1957-58 25 Table 12. Weekly floricultural sales by outlet, 1957-58 26 Table 13. Changes in number of units sold from the preceding week. .27 Table 14. Sales during the six days of sales preceding holidays and special occasions, 1 955-56 28 Table 15. Sales during the six days of sales preceding holidays and special occasions, 1 956-57 29 Table 1 6. Sales during the six days of sales preceding holidays and special occasions, 1 957-58 31 Table 17. Shrinkage as a percent of retail sales 35 Table 1 8. Percent of sales of each kind of plant displayed, by outlet and year 37 Table 19. Markup and margin on floricultural products sold, by price class and year 39 Table 20. Gross retail margin for floricultural products, by outlet and year 41 Table 21 . Weekly gross returns and profits per square foot of display space for floricultural products, by outlet and year 41 FLORICULTURAL SALES IN MASS MARKET OUTLETS By R. A. KELLY' THE SALE OF POTTED PLANTS is of increasing importance to northern greenhouse operators. Their sales of cut flowers are meeting in- creased competition because flowers can be raised outdoors in the south and west at considerably less expense than they can be raised in greenhouses. Potted plants grown in those regions do not offer as much competition because the bulkiness and expense of transporting most potted plants for any distance still give the advantage to pro- ducers close to the market, even though their production costs are higher. Low-cost cut flowers that are adaptable to northern production methods may offer another way of meeting this competition. Despite the increase in competition in the sales of cut flowers, de- creases in total production in the northern states will probably be slow because of investments in facilities that have limited alternative uses. Therefore, any change in the marketing structure which increases the retail sales of potted plants and of inexpensive cut flowers is impor- tant to greenhouse operators. Fortunately, the demand for low-cost plants is increasing as a part of the present emphasis on outdoor and suburban living. The demand for both cut and potted flowers has traditionally been met, for the most part, by floral shops and, to a limited extent, by other outlets. The object of the present study has been, first, to com- pare flower and plant sales at a supermarket and at a variety store with sales at a floral shop and, second, to compare the profits on the items offered in this experiment with the profits on other merchandise at outlets comparable to those studied here. In view of the considerable research that has been done on the marketing problems of other segments of agriculture, it may seem surprising that relatively few resources have been used to analyze the 1 R. A. KELLY, Associate Professor of Agricultural Economics. John R. Culhert, Associate Professor of Floriculture cooperated by rendering technical assistance and by providing flowers, plants, and storage space. The field work was done by Floyd Griffith, Howard Gingerich, and O'Dean Hubbard, graduate assistants in Agricultural Economics. Judith Kingwell, a graduate student in Floriculture, made the floral arrangements. 6 BULLETIN NO. 675 [August, marketing of horticultural specialties. The lack of such research may have been due to a failure among research agencies to recognize the importance of the industry or to its lack of awareness of the importance of the inefficiencies present in its marketing structure. The underlying premise of the research project was that total sales of inexpensive potted plants and bunches of cut flowers could be in- creased if such plants and flowers were offered for sale in mass market outlets. A specific objective was to test consumer acceptance of cut flowers and potted plants offered for sale in different mass market retail outlets. The reasons for limiting the study to small potted plants and small bunches of cut flowers were: 1. These are grown in Illinois and can compete successfully with out-of -state products. 2. The demand for these products is elastic. 3. These products are satisfactory for "everyday use" as well as for holidays and as inexpensive gifts. 4. They are relatively easy to handle in retail outlets. The same stores were used during the sales period each year for three years: a supermarket, a variety store, and a retail floral shop. The scope of the study was limited to the number of variables that could be controlled in the stores. The variables included the retail price, the relative quality of the displays between the outlets, the size of the display area, and the kinds of plants displayed. EXPERIMENTAL PROCEDURE Display period During 1955-56, potted plants were sold for 12 weeks from No- vember 7 to February 11, with the two weeks from December 19 to January 2 off for inventory; and cut flowers were sold for 8 weeks from February 13 to April 14. No flowers were offered for sale in the experiment during the week before Easter, the first season. During the second year, the project was begun on October 10, 1956, and ended May 11, 1957. The plants were removed from the three outlets for the two weeks from December 24, 1956, to January 7, 1957. The third year, the experiment was conducted from December 10 to May 10, with inventory from December 30 to January 7. Each year, the changes in the varieties or kinds of flowers on display were almost always made at the beginning of the week. J961] FLORICULTURAL SALES 7 Methods of procurement During each year the plants were delivered to each of the outlets on consignment. This arrangement was advantageous in that it per- mitted control of the factors being studied. Its principal disadvantage was an occasional lack of cooperation by the store personnel in assist- ing to maintain quality and complete displays, particularly at the mass market outlets. Sources of supply The Department of Horticulture supplied the potted chrysanthe- mums and the bunches of cut chrysanthemums used the first year. During the second year of the study, the project was conducted on a semicommercial basis. A wholesale grower supplied approximately 62 percent of the volume of plants, which included potted chrysanthe- mums, cyclamens, kalanchoes, and poinsettias. Deliveries were made to the horticulture greenhouse once a week, generally on Wednesday, where they were cared for until they were moved into the stores. The floricultural division grew the chrysanthemums that were sold in bunches, the cinerarias, the coleuses, and the primulas during 1957-58. The other plants were purchased from the wholesale grower. Packaging In 1955-56, the chrysanthemum plants were in 3-inch standard clay pots. The clay pots were inserted into 4-inch, green papier mache pots. Each potted plant was placed in a cellophane bag, which was held around the lower part of the pot with a rubber band. The bags were a protection against stem breakage, and they facilitated watering, but the type of bag used permitted condensation, which reduced the attractiveness of the display. The papier mache container and the plastic bag were not used the second and third years. This eliminated the effect of these packaging techniques on the rate of sale and reduced the labor cost. The removal of the plastic bag furnished an opportunity to see how well the plants would stand up without protection. After the first year, 3^-inch stepdown clay pots were used instead of 3-inch pots. The cut chrysanthemums were packaged in cellophane bags each year, and the bags were fastened at the lower part of the stems with rubber bands or paper-covered wire. The size of the bunches was equivalent to six carnations, regardless of the number of chrysanthe- mum stems needed. Consequently, the number of stems in different 8 BULLETIN NO. 675 [Augusf, packages varied from 5 to 12, depending on the variety. With the exceptions of begonias in 3-inch pots and some poinsettias in 5-inch pots, the plants during the last two years were in 3i/£-inch pots. The third year floral arrangements were also sold. They were packaged similarly to chrysanthemums. Displays and equipment The displays were not alike in every respect at all times, but were comparable in area, number of varieties on display, number of plants on display each morning, price per unit, and general quality. In 1955- 56, the bunched flowers were held upright by a wooden and fiberboard box inverted over a watertight pan. The pan was covered with a light chicken wire. The stems were inserted through the holes in the fiber- board, through the chicken wire, and into the water in the pans. How- ever, when the customers examined the flowers, they did not always replace the stems through the wire. Hence the stems frequently were neither in the water nor were they held upright. This problem was eliminated in subsequent years by fastening sheet metal tubes to the underside of the fiberboard rack and eliminating the chicken wire. The papier mache pots used the first year were a part of the display as well as a means of packaging. The plants were kept fresh by putting water in the papier mache pots. Although the papier mache enhanced the appearance of the display, it had four principal disadvantages: (1) it cost more, (2) additional labor was required for insertion, (3) usually the plant had to be removed from the papier mache pot for watering, which increased labor costs, and (4) when the plants were sold, the water had to be removed. Removing the water took time and required a place to dispose of it. In succeeding years the potted plants were displayed in metal or clay pans which needed about 2 inches of water a day for evaporation and plant requirements. This method had the disadvantage of keeping the pots wet most of the time and made them unpleasant for the cus- tomers to handle. However, this procedure did help decrease the time required to care for the plants and increased their display period by reducing wilting. Retail outlets The same supermarket, variety store, and floral shop were used throughout the study, the floral shop being used as a check against the results in the other stores. The supermarket was near the downtown shopping district and had the largest volume of store traffic. Over the three years, its estimated J96J] FLORICULTURAL SALES 9 store traffic averaged 12,500 customers per week, with approximately half the weekly traffic on Friday and Saturday. The variety store was located in the downtown shopping district, the second store from the busiest intersection in town, with an esti- mated weekly store traffic of approximately 9,700 customers for the three years. As in the supermarket, about half of the weekly traffic- was on Friday and Saturday. The retail floral shop was located around the corner from the va- riety store, approximately half a block from the intersection. The walk-in trade accounted for a substantial amount of its business. By the next year, the floral shop had moved next to a supermarket several blocks from the center of town, where the walk-in trade was approxi- mately half that at the old location. By the third year, the florist moved his store to the edge of a new subdevelopment about two miles from the center of town and the walk-in trade was further reduced. However, the floral shop did much of its business by telephone. SALES: MASS MARKETS AND FLORAL SHOP RETAIL SALES Method of collecting data Data on the number on hand of each kind and variety of plant and flower, and on the number sold, delivered, and removed were obtained each morning at each outlet as the flowers were delivered and the displays arranged. Limitations The analysis of the sale of floral products through mass market outlets was complicated by the difficulty of keeping the outlets supplied with complete and comparable displays. Since the quantity on display varied during the day, the available sales data may not have measured the total potential demand for the floricultural products. The analysis is further complicated by the fact that the kinds of plants and flowers displayed were different each year. During the first season, only chrysanthemums were offered for sale. Potted chrysanthe- mums were offered first, then bunches of cut chrysanthemums, but not both at once. The second year, the display was expanded to include asters, begonias, calceolarias, potted chrysanthemums, bunched cut chrysanthemums, cinerarias, coleuses, cyclamens, kalanchoes, poinset- tias, and primulas. The third year, the list was the same except that the asters, kalanchoes, and primulas were replaced with schizanthuses and arrangements of cut flowers. 10 BULLETIN NO. 675 [August, 80 70 60 50 40 30 20 10 UNITS SOLD 1955-56 VARIETY STORE ^- l i l i I l l I I l I I I I I I I I I I i l I 7 14 21 28 5 12 19 26 2 9 16 23 30 6 13 20 27 5 12 19 26 2 9 NOV. DEC. JAN. FEB. MAR. APRIL UNITS SOLD 1956-57 10 15 22 29 5 12 19 26 3 10 17 24 31 7 14 20 28 4 II 18 25 4 II 18 25 I 8 15 22 29 6 OCT. NOV. DEC. JAN. FEB. MAR. APRIL MAY Trends in weekly floricultural sales by outlet, 1955-56 (top), 1956-57 (below), am 1957-58 (opposite page). (Fig. T ?96?] FLORICULTURAL SALES The lists of flowers sold each year and the number of weeks each was on sale are given in Table 17. Tables 6 and 11 show which weeks each kind of flower was on sale during the second and third seasons. Trend in weekly sales In 1955-56, the trend in weekly sales at the supermarket began rising after January 2 and continued upward through the remainder of the period (Fig. 1). In comparison, the sales trend at the other mass market outlet rose slightly until January and then declined. The sales trend at the floral shop increased gradually until about mid- February and then leveled off. The sales trend lines in the supermarket and in the variety store were almost reversed the second year, declining in the former after mid-February and rising steadily in the variety store (Fig. 1). Even though sales at the floral shop were greater during the second season, the trend was quite similar to that of the first year. no 100 90 60 70 60 50 40 30 20 10 - UNITS SOLD 1957-58 VARIETY STORE I I I I I I I I I I I I I I I I I I I 10 16 23 30 7 13 20 27 3 10 17 24 3 10 17 24 31 7 14 21 28 5 DEC. JAN. FEB. MAR. APRIL MAY Fig. 1 — concluded 12 BULLETIN NO. 675 [August, During the third season, the sales trends in the mass market out- lets were nearly identical until the last few weeks, when sales at the supermarket began leveling off. The trend in the floral shop was similar to that of the two preceding years but rose slightly faster. The rising trend at each mass market outlet for 2 years out of 3 and at the floral shop for all three years indicates an increasing con- sumer acceptance during each season for the low-cost type of items offered in this display. Total sales During the three years, 11,535 plants and bunches were sold through the three outlets. Approximately 41 percent of them were sold at the supermarket, 36 percent at the variety store, and 23 percent at the floral shop (Table 1 and Fig. 2). The total number of plants sold through the variety store for the whole period of the experiment was not as great as through the supermarket, but the trend in sales indi- cated that the variety store was an increasingly acceptable outlet for floricultural products, for its share of the yearly total sales rose by 15.5 percent over the three years and its sales the final season were greater than those at the supermarket. Total sales declined 27.5 percent during the third year because of the shorter period that the plants were offered for sale. The decrease at the variety store and at the floral shop was 20 and 17 percent, respectively, but total sales at the supermarket fell nearly 39 percent from the previous year. Even though the floral shop moved locations each year, with each move resulting in less walk-in trade, the sales of the inexpensive cash- and-carry items used in the experiment held up remarkably well, indi- cating a possibility for increased total sales. Average weekly sales Weekly sales averaged 64 units per store in 1956-57 and 1957-58 compared with 32.5 the first year, an increase of nearly 97 percent (Fig. 3). Of the three outlets, the variety store had the greatest in- crease in volume between the first and second year, 191 percent com- pared with an increase of 86 percent at the supermarket and 35 percent at the floral shop. Average weekly sales further increased at the variety store and floral shop the third year, but the average weekly volume at the supermarket declined 15.5 percent from the weekly average for the second year. J96J] FLORICULTURAL SALES Table 1. — Floricultural Sales by Outlet and Year 13 Supermarket Variety store Floral shop Total Year Num- Per- Num- Per- Num- Per- Num- Per- ber cent ber cent ber cent ber cent 1955-56... 889 45.6 478 24.5 584 29.9 1,951 100.0 1956-57... . 2,395 43.1 2,017 36.3 1,145 20.6 5,557 100.0 1957-58... . 1 ,466 36.4 1,610 40.0 951 23.6 4,027 100.0 Total . . . . 4,750 41.2 4,105 35.6 2,680 23.2 11,535 100.0 2,500 2,000 =J 1,500 1,000 500 I UNITS PERCENT 1955-56 1956-57 1957-58 SUPERMARKET 1955-56 1956-57 1957-58 VARIETY STORE 1955-56 1956-57 1957-58 FLORAL SHOP 100 SO 60 20 Total number of sales and percent, by outlet and year. (Fig. 2) VARIATION IN WEEKLY SALES AT EACH OUTLET Limitations Important limitations to the analysis of changes in weekly sales include the base period used and the many variables which could not be controlled. The change in the number of plants sold from the aver- age of the previous two weeks was used as a measure of the result of the introduction of the various factors considered and of their effects on floricultural sales. The adequacy of this base is questionable, yet it 14 BULLETIN NO. 675 [August, would seem that the factors would show their effects over a short period of time. Also, no measures are available to separate out the effects of the individual variables when two or more occurred simul- taneously. For instance, the introduction of a new kind of plant, the occurrence of a holiday, and the addition of competing floricultural products by the stores may all have happened at the same time, and the 1955-56 1956-57 1957-58 AVERAGE 1955-56 1956-57 1957- 58 AVERAGE 1955-56 1956-57 1957-58 AVERAGE 1955-56 1956-57 1957-58 AVERAGE SUPERMARKET VARIETY STORE FLORAL SHOP I AVERAGE, ALL STORES J^ I 1 I I 1 0 10 20 30 40 50 UNITS SOLD Weekly average sales, by outlet and year. 60 70 80 (Fig. 3) 1967] FLORICULTURAL SALES 31 Table 16. — Sales During the Six Days of Sales Preceding Holidays and Special Occasions Compared With the Average Weekly Sale for the Previous 12 Sales Days, 1957-58 Holidays and special occasions Number sold during the six days of sales preceding a holiday or spe- cial occasion Average weekly sale, pre- vious 12 sales days Difference in number Percent change Supermarket Christmas' 38 33.0 5.0 15.2 Valentine's Day 75 63.0 12.0 19.0 Easter 104 113.5 -9.5 -8.4 Mother's Day 102 82 . 5 19.5 23 . 6 Total 325 292.0 33.0 11.3 Variety Store Christmas" 26 17.0 9.0 52.9 Valentine's Day 78 63.5 14.5 22.8 Easter 112 108.5 3.5 3.2 Mother's Day 115 99.0 16.0 16.2 Total 337 288.0 49.0 17.0 Florist Christmas' 38 9.0 29.0 322.2 Valentine's Day 41 40.0 1.0 2.5 Easter 77 69.0 8.0 11.6 Mother's Day 57 77.5 -20.5 -26.5 Total 214 195.5 18.5 9.5 Total Christmas" 102 57.0 43.0 72.9 Valentine's Day 194 166.5 27.5 16.5 Easter 293 291.0 2.0 .7 Mother's Day 274 259.0 15.0 5.8 Total 876 775.5 100.5 13.0 • Omitting potted chrysanthemums, which were introduced the week before Christmas. sale. During the same period in 1957-58, only begonias and potted and bunched chrysanthemums were displayed. However, the kinds of flowers sold before Easter were practically the same for both seasons. Competing products and displays Competing products are denned as other plants and flowers that the retail outlets offered for sale in the area in which the flowers offered in this experiment were displayed. These flowers helped at- tract attention to the display area, but they also directly competed for 32 BULLETIN NO. 675 [August, the customers' preference. The analysis of the effect of the competi- tion of these items is limited by the fact that most of the competing products were added during the holiday seasons, when the sales of the items offered in the experiment usually increased regardless of the presence of other factors. For short periods in 1955-56, gladioli were added at the super- market in January and jonquils in February, March, and April. The variety store sold daffodils for a week in March. Sales of the plants in the experiment rose 13 percent at the supermarket and fell 21 per- cent at the variety store. In 1956-57, coleuses, African violets, daffodils, philodendrons, rose bushes, pansies, and 2-inch chrysanthemums were displayed all or parts of 10 weeks at the supermarket. Sales of the plants that were a part of the experiment declined from the preceding week in 6 of those weeks and increased in 4. Sales declined 2 weeks in the variety store and rose in 3 when competitive azaleas, daffodils, and jonquils were added, the change ranging from a 56-percent decline to a 75-percent increase. The important changes in competition at the floral shop during 1956-57 occurred at the various holidays when large poinsettias and jonquils were displayed. Sales of the items in the experiment generally increased at all three outlets, even when the additional com- petitive displays were present. It is not possible to tell, by an analysis of the available data, how much effect the holidays and special occa- sions had on sales and how much effect the competition had when they occurred at the same time. In 1957-58, the results were similar to those of the preceding year. Sales of competitive products had no consistent, continuous effect on sales of the plants and flowers in the experiment. The unevenness of the effect of the competition from other plants probably was due to differences in the kinds of plants and in their sizes and prices. Changes in display location Changes in the location of the displays were not planned as mer- chandising techniques for the experiment but resulted from variations in the space allocated to the project by the stores. Relatively few of the weeks during which there was a change in the location of the display were associated with changes in total weekly sales. For the two weeks prior to Christmas the first year at the variety store, the items were placed on racks 6 feet tall. This did not permit customers to get a sufficient view to stimulate impulse buying. Sales decreased by 70 percent but increased when the display was moved back to the table. J96J] FLORICULTURAl SALES 33 At the supermarket the same year, some of the plants were placed atop racks displaying nonfood items. These racks were new equip- ment, and it was felt that customers regarded the plants as a part of the store decoration because sales from that display were practically nil. An attractive display was made for a short period at the florist's by displaying the plants in a window with the price plainly marked. This display was not directly associated with increased weekly sales, but it seems probable that the display attracted the attention of passers-by and affected the volume of the walk-in trade of the store. Addition of various kinds of plants to the display During 1955-56, the change from potted to bunched chrysanthe- mums had no appreciable effect on sales. In 1956-57, the introduction of poinsettias preceding the Christmas holiday and the reintroduction of cinerarias preceding Easter contributed to the increase in sales. The introduction of cinerarias and the reintroduction of calceolarias in two stores in weeks following holidays were both associated with decreased sales. However, the sales of the calceolarias at the variety store were as high when these flowers were reintroduced as they had been be- fore. Since sales of all other flowers slumped, thus causing the general decline in sales, it appears that the reintroduction of these plants helped somewhat to counteract the slump. Not counting the weeks when special occasions caused sales to increase, the introduction of new plants increased sales over the previous week for five of the remaining six weeks. It appears that sales were generally larger when new kinds of plants were added to the display. The value of the new introductions as a merchandising technique is difficult to determine because of the large variations in weekly sales and because of the substitution effect of the newly introduced flowers for those already on display. For example, during the 1956-57 season, the influence of the approaching holiday and the introduction of kalanchoes three weeks before Christ- mas resulted in an over-all increase in sales of 104 percent from the previous week. The kalanchoe sales were larger than the increase in weekly sales, indicating that the new kind of plant was a stimulus to total sales but that it had a detrimental effect on the movement of other plants. Weekly sales of each kind of plant displayed, with the exception of coleuses and bunched chrysanthemums, generally tended to decrease during the period it was on display. This indicated that introducing 34 BULLETIN NO. 675 [August, new kinds of flowers at various time intervals would be a merchan- dising technique that would increase total sales. The most advantage- ous interval to use would vary with the kinds of plants used and according to the season of the year. During 1957-58, the introduction of new varieties of a given kind of plant frequently had no appreciable effect on sales, largely because of the small number of units of each variety displayed at each store. Shrinkage By shrinkage is meant removal of unsold plants from the display. The term applies to plants removed to make room for the display of new kinds of plants and to plants removed whenever the display was concluded, as well as to the plants removed because they had gone out of condition. Plants which were out of condition usually had suffered wilting, loss of leaves, age of blossom, or stem breakage. The com- pactness and the self-service aspect of the display encouraged customer handling, and this contributed to the plants going out of condition. Daily delivery and removal of some plants were made to keep the quality of the displays high and comparable between the stores. Some plants were probably removed sooner than they would have been in normal retail practice. The limited display space allotted to the project necessitated removing a few plants that were still in salable condition when new kinds were introduced. Data on these plants were not separated from the data on those removed when out of condition. The price was constant over extended periods, and no special mark-down procedure was used (as is done with some products in mass markets) to help sell those plants that had been on display for long periods, that had broken stems, that had lost part of the leaves, or that were for particular holidays, such as poinsettias. Neither were these flowers used in corsages, arrangements, or bouquets as is often done with slightly out-of-condition flowers at floral shops. The slow sale of various kinds of the floricultural division's experi- mental varieties helped increase the number removed. Some of these plants either were not readily accepted by the customers or were not able to withstand the in-store handling. The average total shrinkage varied from 25.5 to more than 32 percent during the three years (Table 17). In 1955-56, the variety store had a larger shrinkage than the other outlets due, in part, to the difficulty in keeping the plants properly watered. The evaporation rate was particularly high, and the store personnel assigned to the I96J] FLORICULTURAL SALES 35 cfl "3 a -2-i u. <" 1 * .Si u "s 3 O !« V. «5 u ^ ^1 ]| U, * 1 .K jy 3 O •*^ 4) C 'C 0 £« 3" rt C/3 £ o * E *> -s c i) O = £ 8 — -10 fNr-) — i— oooooovo t— — oo 10 Ov Tj< -H fN Tf '— '-' CNt— «-i 10 oo en £• = 3 § E E-S« •s-S£ - c H O 3 i- a in C in C 3 « 5 P m £ 3 .2 3 S o p-w £ § §"S-g 2 S « | 9! = 0 So^S e *iji~^ c £ ' ' o 3 . o 36 BULLETIN NO. 675 [August, project occasionally neglected to add water during the day. During the second year, a change in the type of display equipment made it possible to keep the plants watered without having to rely so heavily