where:
 DH=dining room area (in hundreds of square feet) times
 hours the dining room is open per week,
 BH=bar room area (in hundreds of square feet) times hours
 the bar is open per week.

 DATA COLLECTION
 The study incorporated primary data collected via question-
 naires from individual businesses on all variables other than
 price (r) and quantity (IW) [and price of room (PR) in the
 motel-hotel category]. Price and quantity data were collected
 directly from 17 water companies in Dade county and the Florida
 Keys [15, p. 27]. The companies included in the sample are
 shown in Table 2. A total of 811 businesses were mailed question-
 naires and 308 were returned for a 37 percent return rate. Us-
 able questionnaires gave 103 data points in addition to the 93
 obtained from DHR for the hotels-motels group "secondary"
 model.

 Table 2.-Water companies included in sampling process.
 Miami Dade Water and General (GEN) Water Works
 Sewage Authority City of Coral Gables
 Hialeah Water and Sewage (part of GEN)
 City of North Miami City of Homestead
 North Miami Beach South Miami (part of GEN)
 City of West Miami Florida City
 Coral City Utilities Key Biscayne (GEN)
 Miami Springs Water & Sewage City of Opa Locka
 Florida Utilities Florida Keys
 City of Miami Beach


 RESULTS FOR INDIVIDUAL STORES
 Several mathematical equations were estimated relating the
 dependent and the independent variables discussed previously.
 We discuss each group in turn.

 Department Store Group
 The equation estimated for the department stores is repre-
 sented by Equation (1.1) in Table 3. The price (r) and area (A)
 variables were found to be significant at the 0.01 probability level.
 The restaurant variable (IRA) was significant at only the 0.20
 probability level. The coetficient of multiple determination was
 estimated at R-=0.82 and the adjusted was -'-=0.78, indicating