Stewart Calculus ET 5e 0534393217;14. Partial Derivatives; 14.

8 Lagrange Multipliers

1. At the extreme values of f , the level curves of f just touch the curve g(x,y)=8 with a common
tangent line. (See Figure 1 and the accompanying discussion.) We can observe several such
occurrences on the contour map, but the level curve f (x,y)=c with the largest value of c which still
intersects the curve g(x,y)=8 is approximately c=59 , and the smallest value of c corresponding to a
level curve which intersects g(x,y)=8 appears to be c=30 . Thus we estimate the maximum value of f
subject to the constraint g(x,y)=8 to be about 59 and the minimum to be 30 .

2. (a) The values c= 1 and c=1.25 seem to give curves which are tangent to the circle. These values
2 2 2
represent possible extreme values of the function x +y subject to the constraint x +y =1 .

1
(b)  f = 2x,1 ,   g= 2 x,2 y . So 2x=2 x either  =1 or x=0 . If  =1 , then y= and so
2
3
x= (from the constraint). If x=0 , then y= 1 . Therefore f has possible extreme values at the
2
3 1 3 1 5
points ( 0, 1 ) and  , . We calculate f  , = (the maximum value),
2 2 2 2 4
f (0,1)=1 , and f (0, 1)= 1 (the minimum value). These are our answers from (a).

2 2 2 2
3. f (x,y)=x  y , g(x,y)=x +y =1  f = 2x, 2y ,   g= 2 x,2 y . Then 2x=2 x implies x=0 or
2 2
 =1 . If x=0 , then x +y =1 implies y= 1 and if  =1 , then  2y=2 y implies y=0 and thus x= 1 .
Thus the possible points for the extreme values of f are (  1,0 ) , ( 0, 1 ) . But f (  1,0 ) =1 while
2 2
f (0, 1)= 1 so the maximum value of f on x +y =1 is f ( 1,0)=1 and the minimum value is
f (0, 1)= 1 .

2 2
4. f (x,y)=4x+6y , g(x,y)=x +y =13  f = 4,6 ,   g= 2 x,2 y . Then 2 x=4 and 2 y=6 imply
2 3 2 3 2 2
2 2 13
x= and y= . But 13=x +y = +  13=   = 1 , so f has possible extreme
    2

values at the points ( 2,3) , (  2, 3) . We compute f (2,3)=26 and f ( 2, 3)= 26 , so the maximum
2 2
value of f on x +y =13 is f (2,3)=26 and the minimum value is f ( 2, 3)= 26 .
1
Stewart Calculus ET 5e 0534393217;14. Partial Derivatives; 14.8 Lagrange Multipliers

2 2 2 2
5. f (x,y)=x y , g(x,y)=x +2y =6  f = 2xy,x ,   g= 2 x,4 y . Then 2xy=2 x implies x=0 or
2
 =y . If x=0 , then x =4 y implies  =0 or y=0 . However, if y=0 then g(x,y)=0, a contradiction. So
2 2 2
 =0 and then g(x,y)=6 y= 3 . If  =y , then x =4 y implies x =4y , and so g(x,y)=6
2 2 2
4y +2y =6 y =1 y= 1 . Thus f has possible extreme values at the points ( 0, 3 ) , (  2,1 ) ,
and (  2, 1 ) . After evaluating f at these points, we find the maximum value to be f ( 2,1)=4 and
the minimum to be f ( 2, 1)= 4 .

2 2 4 4 3 3 3
6. f (x,y)=x +y , g(x,y)=x +y =1  f = 2x,2y ,   g= 4 x ,4 y . Then x=2 x implies x=0 or
1 4 4 3 1
 = . If x=0 , then x +y =1 implies y= 1 . But y=2 y implies y=0 so x= 1 or  = and
2 2
2x 2y
2 2 4 1
x =y and 2x =1 so x= 4 . Hence the possible points are ( 0, 1 ) , (  1,0 ) ,
2
1 1 4 4
 , 4 , with the maximum value of f on x +y =1 being
[4]2 2
1 1 2
f  4 , 4 = = 2 and the minimum value being f (0, 1)= f ( 1,0)=1 .
2 2 2
2 2 2
7. f (x,y,z)=2x+6y+10z , g(x,y,z)=x +y +z =35  f = 2,6,10 ,   g= 2 x,2 y,2 z . Then 2 x=2
1 3 5 2 2 2 1 2 3 2 5 2
, 2 y=6 , 2 z=10 imply x= , y= , and z= . But 35=x +y +z = + + 
     
35
35=   = 1 , so f has possible extreme values
2

2 2 2
at the points ( 1,3,5) , (  1, 3, 5) . The maximum value of f on x +y +z =35 is f (1,3,5)=70 , and the
minimum is f ( 1, 3, 5)= 70 .

2 2 2
8. f (x,y,z)=8x 4z , g(x,y,z)=x +10y +z =5  f = 8,0, 4 ,   g= 2 x,20 y,2 z . Then 2 x=8 ,
4 2 2 2 2 4 2 2 2 2
20 y=0 , 2 z= 4 imply x= , y=0 , and z= . But 5=x +10y +z = +10 ( 0 ) +  
   
20
5=   = 2 , so f has possible extreme values at the points ( 2,0, 1 ) , (  2,0,1 ) . The maximum of
2

2 2 2
f on x +10y +z =5 is f (2,0, 1)=20 , and the minimum is f ( 2,0,1)= 20 .
2 2 2
9. f (x,y,z)=xyz , g(x,y,z)=x +2y +3z =6 
f = yz,xz,xy ,   g= 2 x,4 y,6 z . Then  f =  g
2 2 2 2 2 2 2 2 2
implies  =(yz)/(2x)=(xz)/(4y)=(xy)/(6z) or x =2y and z = y . Thus x +2y +3z =6 implies 6y =6 or
3
2
Stewart Calculus ET 5e 0534393217;14. Partial Derivatives; 14.8 Lagrange Multipliers

2 2 2
y= 1 . Then the possible points are 2 , 1, , 2 , 1, ,  2 , 1,
3 3 3
2 2
,  2 , 1, . The maximum value of f on the ellipsoid is
, occurring when all
3 3
2
coordinates are positive or exactly two are negative and the minimum is  occurring when 1 or 3
3
of the coordinates are negative.

2 2 2 2 2 2 2 2 2 2 2 2
10. f (x,y,z)=x y z , g(x,y,z)=x +y +z =1  f = 2xy z ,2yx z ,2zx y ,   g= 2 x,2 y,2 z .
2 2 2 2 2 2
Then  f =  g implies (1)  =y z =x z =x y and   0 , or (2)  =0 and one or two (but not three) of
2 2 2 1
the coordinates are 0 . If (1) then x =y =z = . The minimum value of f on the sphere occurs in case
3
1
(2) with a value of 0 and the maximum value is which arises from all the points from (1), that is,
27
1 1 1 1 1 1 1 1 1
the points  , , ,  , , ,  , , .
3 3 3 3 3 3 3 3 3
2 2 2 4 4 4
11. f (x,y,z)=x +y +z , g(x,y,z)=x +y +z =1
3 3 3
 f = 2x,2y,2z ,   g= 4 x ,4 y ,4 z .
Case 1: If x 0 , y 0 and z 0 , then  f =  g implies  =1/ 2x =1/ 2y =1/ 2z or x =y =z ( 2) ( 2) ( 2) 2 2 2

4 1 1 1 1 1 1 1
and 3x =1 or x= 4 giving the points  4 , 4 , 4 ,  4 , 4 , 4 ,
3 3 3 3 3 3 3
1 1 1 1 1 1
 4 , 4 , 4 ,  4 , 4 , 4 all with an f  value of 3 . Case 2: If one of
3 3 3 3 3 3
the variables equals zero and the other two are not zero, then the squares of the two nonzero
1
coordinates are equal with common value and corresponding f value of 2 . Case 3: If exactly
2
two of the variables are zero, then the third variable has value  1 with the corresponding f value of 1
4 4 4
. Thus on x +y +z =1 , the maximum value of f is 3 and the minimum value is 1 .
4 4 4 2 2 2
12. f (x,y,z)=x +y +z , g(x,y,z)=x +y +z =1
3 3 3
 f = 4x ,4y ,4z ,   g= 2 x,2 y,2 z .
2 2 2 2 2 2 1
Case 1: If x 0 , y 0 and z 0 then  f =  g implies  =2x =2y =2z or x =y =z = yielding 8
3
1
points each with an f  value of .
3
Case 2:
3
Stewart Calculus ET 5e 0534393217;14. Partial Derivatives; 14.8 Lagrange Multipliers

If one of the variables is 0 and the other two are not, then the squares of the two nonzero coordinates
1 1
are equal with common value and the corresponding f  value is .
2 2
Case 3: If exactly two of the variables are 0 , then the third variable has value  1 with corresponding
2 2 2 1
f  value of 1 . Thus on x +y +z =1 , the maximum value of f is 1 and the minimum value is .
3

2 2 2 2
13. f (x,y,z,t)=x+y+z+t , g(x,y,z,t)=x +y +z +t =1 1,1,1,1 = 2 x,2 y,2 z,2 t , so
2 2 2 2
 =1/(2x)=1/(2y)=1/(2z)=1/(2t) and x=y=z=t . But x +y +z +t =1 , so the possible points are
1 1 1 1 1 1 1 1
 , , , . Thus the maximum value of f is f , , , =2 and the minimum
2 2 2 2 2 2 2 2
1 1 1 1
value is f  , , , = 2 .
2 2 2 2
2 2 2
14. f x ,x ,... ,x =x +x +   +x , g x ,x ,... ,x =x +x +   +x =1
( ) ( )
1 2 n 1 2 n 1 2 n 1 2 n
1,1,... ,1 = 2 x ,2 x ,... ,2 x , so  =1/ 2x =1/ 2x =   =1/ 2x
( ) ( ) ( ) and x =x =   =x . But
1 2 n 1 2 n 1 2 n
2 2 2
x +x +   +x =1 , so x = 1/ n for i=1 , ... , n . Thus the maximum value of f is
1 2 n i
f ( 1/ n,1/ n,... ,1/ n ) = n and the minimum value is f (  1/ n, 1/ n,... , 1/ n ) = n .
2 2
15. f (x,y,z)=x+2y , g(x,y,z)=x+y+z=1 , h(x,y,z)=y +z =4 f = 1,2,0 ,   g=  , , and 
1
  h= 0,2 y,2 z . Then 1= , 2= +2 y and 0= +2 z so  y= =  z or y=1/(2 ) , z= 1/(2 ) .
2
2 2 1
Thus x+y+z=1 implies x=1 and y +z =4 implies  = . Then the possible points are
2 2
( 1, 2 , 2 ) and the maximum value is f ( 1, 2 , 2 ) =1+2 2 and the minimum value is
f ( 1, 2 , 2 ) =1 2 2 .
2 2
16. f (x,y,z)=3x y 3z , g(x,y,z)=x+y z=0 , h(x,y,z)=x +2z =1  f = 3, 1, 3 ,   g=  , ,  ,
  h= ( 2 x,0,4 z ) . Then 3= +2 x ,  1= and  3=  +4 z , so  = 1 ,  z= 1 ,  x=2 . Thus
4 1 1 2
h(x,y,z)=1 implies +2 =1 or  = 6 , so z= ; x= ; and g(x,y,z)=0 implies
2 2 6 6
 
3 6 6 6
y= . Hence the maximum of f subject to the constraints is f , , =2 6 and
6 3 2 6
6 6 6
the minimum is f  , , = 2 6 .
3 2 6

2 2
17. f (x,y,z)=yz+xy,g(x,y,z)=xy=1 , h(x,y,z)=y +z =1  f = y,x+z,y ,   g=  y, x,0 ,

4
Stewart Calculus ET 5e 0534393217;14. Partial Derivatives; 14.8 Lagrange Multipliers

  h= 0,2 y,2 z . Then y= y implies  =1 , x+z= x+2 y and y=2 z . Thus  =z/(2y)=y/(2y) or
2 2 2 2 1 1
y =z , and so y +z =1 implies y= , z= . Then xy=1 implies x= 2 and the possible
2 2
1 1 1 1
points are  2 , , ,  2 , , . Hence the maximum of f subject to the
2 2 2 2
1 1 3 1 1 1
constraints is f  2 , , = and the minimum is f  2 , , = .
2 2 2 2 2 2
Note: Since xy=1 is one of the constraints we could have solved the problem by solving f (y,z)=yz+1
2 2
subject to y +z =1 .

2 2
18. f (x,y)=2x +3y  4x 5  f = 4x 4,6y = 0,0  x=1 , y=0 . Thus ( 1,0 ) is the only critical point
2 2 2 2
of f , and it lies in the region x +y <16 . On the boundary, g(x,y)=x +y =16   g= 2 x,2 y , so
6y=2 y either y=0 or  =3 . If y=0 , then x= 4 ; if  =3 , then 4x 4=2 x x= 2 and y= 2 3 .
Now f (1,0)= 7 , f (4,0)=11 , f ( 4,0)=43 , and f (  2, 2 3 ) =47 . Thus the maximum value of f (x,y)
2 2
on the disk x +y
16 is f (  2, 2 3 ) =47 , and the minimum value is f (1,0)= 7 .

 xy  xy  xy
19. f (x,y)=e . For the interior of the region, we find the critical points: f = ye , f = xe ,
x y
so the only critical point is ( 0,0 ) , and f (0,0)=1 . For the boundary, we use Lagrange multipliers.
2 2  xy  xy
g(x,y)=x +4y =1   g= 2 x,8 y , so setting  f =  g we get  ye =2 x and  xe =8 y . The
 xy 2 2
first of these gives e = 2 x/y , and then the second gives  x( 2 x/y)=8 y x =4y . Solving this
2 2 1 1
last equation with the constraint x +4y =1 gives x= and y= . Now
2 2 2
1 1 1/4 1 1  1/4
f  , =e  1.284 and f  , =e  0.779 . The former are the
2 2 2 2 2 2
maxima on the region and the latter are the minima.

20. (a) The graphs of f (x,y)=3.7 and f (x,y)=350 seem to be tangent

to the circle, and so 3.7 and 350 are the approximate minimum and
2 2
maximum values of the function f (x,y) subject to the constraint (x 3) +(y 3) =9 .
2 2 2 2
(b) Let g(x,y)=(x 3) +(y 3) . We calculate f (x,y)=3x +3y , f ( x,y ) =3y +3x , g (x,y)=2x 6 , and
x y x
g (x,y)=2y 6 ,
y
2 2
and use a CAS to search for solutions to the equations g(x,y)=(x 3) +(y 3) =9 , f = g , and f = g
x x y y
3 3
.The solutions are (x,y)= 3 2 ,3 2  ( 0.879,0.879 ) and
2 2
5
Stewart Calculus ET 5e 0534393217;14. Partial Derivatives; 14.8 Lagrange Multipliers

3 3
( x,y ) = 3+ 2 ,3+ 2  ( 5.121,5.121 ) . These give
2 2
3 3 351 243
f 3 2 ,3 2 =  2  3.673 and
2 2 2 2
3 3 351 243
f 3+ 2 ,3+ 2 = + 2  347.33 , in accordance with part (a).
2 2 2 2

 1    1 1   
21. P(L,K)=bL K , g(L,K)=mL+nK= p  P=  bL K ,(1  )bL K ,   g=  m, n .
1   1  
Then  b(K/L) = m and (1  )b(L/K) = n and mL+nK= p , so  b(K/L) /m=(1  )b(L/K) /n or
 1 
n / m(1  ) =(L/K) (L/K) or L=Kn /[m(1  )] . Substituting into mL+nK= p gives K=(1  ) p/n
and L= p/m for the maximum production.
 1    1 1   
22. C(L,K)=mL+nK , g(L,K)=bL K =Q  C= m,n ,   g=   bL K , (1  )bL K .
m L 1  n K 
Then = and
 b K (1  )b L
 1  n L 1  L  Kn Kn  1 
bL K =Q =  L= and so b K =Q .
m(1  ) K K m(1  ) m(1  )
    1   1 1  1 
Q Qm (1  ) Qm (1  ) Qn 
Hence K= = and L= =
     1   1 1  1 
b ( n / m(1  ) ) bn  bn  bm (1  )
minimizes cost.

23. Let the sides of the rectangle be x and y . Then f (x,y)=xy , g(x,y)=2x+2y= p  f (x,y)= y,x ,
1 1
  g= 2 ,2 . Then  = y= x implies x=y and the rectangle with maximum area is a square with
2 2
1
side length p .
4

24. Let f (x,y,z)=s(s x)(s y)(s z) , g(x,y,z)=x+y+z . Then  f =  s(s y)(s z), s(s x)(s z), s(s x)(s y)
,   g=  , , . \ Thus (1) (s y)(s z)=(s x)(s z) and (2) (s x)(s z)=(s x)(s y) . (1) implies x=y while
(2) implies y=z , so x=y=z= p/3 and the triangle with maximum area is equilateral.

6
Stewart Calculus ET 5e 0534393217;14. Partial Derivatives; 14.8 Lagrange Multipliers

2 2 2 2
25. Let f (x,y,z)=d =(x 2) +(y 1) +(z+1) , then we want to minimize f subject to the constraint
g(x,y,z)=x+y z=1 .  f =  g 2(x 2),2(y 1),2(z+1) = 1,1, 1 , so x=( +4)/2 , y=( +2)/2 ,
 +4  +2  +2
z= ( +2)/2 . Substituting into the constraint equation gives + + =1 3 +8=2  = 2 ,
2 2 2
so x=1 , y=0 , and z=0 . This must correspond to a minimum, so the shortest distance is
2 2 2
d= (1 2) +(0 1) +(0+1) = 3 .

2 2 2 2
26. Let f (x,y,z)=d =(x 1) +(y 2) +(z 3) , then we want to minimize f subject to the constraint
g(x,y,z)=x y+z=4 .  f =  g 2(x 1),2(y 2),2(z 3) = 1, 1,1 , so x=( +2)/2 , y=(4  )/2 ,
 +2 4   +6 4 5
z=( +6)/2 . Substituting into the constraint equation gives  + =4  = , so x= ,
2 2 2 3 3
4 11
y= , and z= . This must correspond to a minimum, so the point on the plane closest to the point
3 3
5 4 11
( 1,2,3) is , , .
3 3 3

2 2 2 2
27. f (x,y,z)=x +y +z , g(x,y,z)=z  xy 1=0  f = 2x,2y,2z =  g=   y,  x,2 z . Then 2z=2 z
implies z=0 or  =1 . If z=0 then g(x,y,z)=1 implies xy= 1 or x= 1/y . Thus 2x=  y and 2y=  x
2 2
imply  =2/y =2y or y= 1 , x= 1 . If  =1 , then 2x= y and 2y= x imply x=y=0 , so z= 1 . Hence
the possible points are (  1, 1,0 ) , ( 0,0, 1 ) and the minimum value of f is f (0,0, 1)=1 , so the
points closest to the origin are ( 0,0, 1 ) .

2 2 2 2 2 2 2 2 2
28. f (x,y,z)=x +y +z , g(x,y,z)=x y z=1  f = 2x,2y,2z =  g= 2 xy z,2 x yz, x y . Then
2 2 2 2 2 2
 y z=1 ,  x z=1 and  x y =2z so y z=x z and x= y . Also 2z/1= x y /  x z so 2z =y and
2 2
( 2
) 2 2

2 2 5 1/10 1/10  2/5

y= 2 z . But x y z=1 implies z>0 and 4z =1 . Thus the points are ( 2 , 2 ,2 ) , and the
minimum distance is attained at each of these.

29. f (x,y,z)=xyz , g(x,y,z)=x+y+z=100  f = yz,xz,xy =  g=  , , . Then  =yz=xz=xy implies

100
x=y=z= .
3

a b c
30. f (x,y,z)=x y z , g(x,y,z)=x+y+z=100
a 1 b c a b 1 c a b c 1 a 1 b c a b 1 c a b c 1
 f = ax y z ,bx y
z ,cx y z =  g=  , , . Then  =ax y z =bx y z =cx y z or
ay cy ay cy 100b 100a
ayz=bxz=cxy . Thus x= , z= , and +y+ =100 implies that y= , x= and
b b b b a+b+c a+b+c

7
Stewart Calculus ET 5e 0534393217;14. Partial Derivatives; 14.8 Lagrange Multipliers

100c
z= gives the maximum.
a+b+c

2 2 2
31. If the dimensions are 2x , 2y and 2z , then f (x,y,z)=8xyz and g(x,y,z)=9x +36y +4z =36
2 2
 f = 8yz,8xz,8xy =  g= 18 x,72 y,8 z . Thus 18 x=8yz , 72 y=8xz , 8 z=8xy so x =4y ,
2 2 2 2 2 1
z =9y and 36y +36y +36y =36 or y= ( y>0 ). Thus the volume of the largest such box is
3
1 2 3
8 =16 3 .
3 3 3
2 2 2 2 2 2 2 2 2 2 2
32. f (x,y,z)=8xyz , g(x,y,z)=b c x+a c y +a b z =a b c 
2 2 2 2 2 2 2 2 2 2
 f = 8yz,8xz,8xy =  g= 2 b c x,2 a c y,2 a b z . Then 4yz= b c x , 4xz= a c y ,
2 2 4yz 4xz 4xy y x z y ay cy
4xy= a b z imply  = 2 2 = 2 2 = 2 2 or 2 = 2 and 2 = 2 . Thus x= , z= ,
b b
bc x ac y abz b x a y c y bz
2 2 2 2 2 2 2 2 2 2 2 2 b a c 8
and a c y +c a y +a c y =a b c , or y= , x= , z= and the volume is abc .
3 3 3 3 3

1 1
33. f (x,y,z)=xyz , g(x,y,z)=x+2y+3z=6  f = yz,xz,xy =  g=  ,2 ,3 . Then  =yz= xz= xy
2 3
2 2 4
implies x=2y , z= y . But 2y+2y+2y=6 so y=1 , x=2 , z= and the volume is V = .
3 3 3

34. f (x,y,z)=xyz , g(x,y,z)=xy+yz+xz=32  f = yz,xz,xy =  g=  (y+z), (x+z), (x+y) . Then (1)

 (y+z)=yz , (2)  (x+z)=xz and (3)  (x+y)=xy . And (1) minus (2) implies  (y x)=z(y x) so x=y or  =z
. If  =z , then (1) implies z(y+z)=yz or z=0 which is false. Thus x=y . Similarly (2) minus (3) implies
2 8
 (z y)=x(z y) so y=z or  =x . As above,   x , so x=y=z and 3x =32 or x=y=z= cm.
6
35. f (x,y,z)=xyz , g(x,y,z)=4(x+y+z)=c  f = yz,xz,xy ,   g= 4 ,4 ,4 . Thus 4 =yz=xz=xy or
1
x=y=z= c are the dimensions giving the maximum volume.
12

36. C ( x,y,z ) =5xy+2xz+2yz , g ( x,y,z ) =xyz=V 

 C= 5y+2z,5x+2z,2x+2y =  g=  yz, xz, xy . Then (1)  yz=5y+2z , (2)  xz=5x+2z , (3)
 xy=2 ( x+y ) and (4) xyz=V . Now (1)  (2) implies  z(y x)=5(y x) , so x=y or  =5/z , but z can’t be 0
, so x=y . Then twice (2) minus five times (3) together with x=y implies  y(2x 5y)=2(2z 5y) which
5 5 3 3 2
gives z= y . Hence y =V and the dimensions which minimize cost are x=y= V units,
2 2 5

8
Stewart Calculus ET 5e 0534393217;14. Partial Derivatives; 14.8 Lagrange Multipliers

1/3 5 2/3
z=V units.
2

37. If the dimensions of the box are given by x , y , and z , then we need to find the maximum value
2 2 2 2 2 2 2
of f (x,y,z)=xyz ( x,y,z>0 ) subject to the constraint L= x +y +z or g(x,y,z)=x +y +z =L .
yz xz xy
 f =  g yz,xz,xy = 2x,2y,2z , so yz=2 x  = , xz=2 y  = , and xy=2 z  =
2x 2y 2z
yz xz 2 2 yz xy
. Thus  = =  x =y  x=y and  = =  x=z
2x 2y 2x 2z
2 2 2 2 2 2
. Substituting into the constraint equation gives x +x +x =L  x =L /3 x=L/ 3 =y=z and the
3 3
maximum volume is ( L/ 3 ) =L / ( 3 3 ) .

38. Let the dimensions of the box be x , y , and z , so its volume is f (x,y,z)=xyz , its surface area is
g(x,y,z)=xy+yz+xz=750 and its total edge length is h(x,y,z)=x+y+z=50 . Then
 f = yz,xz,xy =  g+  h=  (y+z), (x+z), (x+y) +  , , . So (1) yz= (y+z)+ , (2)
50
xz= (x+z)+ , and (3) xy= (x+y)+ . Notice that the box can’t be a cube or else x=y=z= but then
3
2500
xy+yz+xz=  750 . Assume x is the distinct side, that is, x y , x z . Then (1) minus (2) implies
3
z ( y x ) = (y x) or  =z , and (1) minus (3) implies y(z x)= (z x) or  =y . So y=z= and x+y+z=50
2
implies x=50 2 ; also xy+yz+xz=750 implies x(2 )+ =750 . Hence
2
750  2 50 5 10
50 2 = or 3  100 +750=0 and  = , giving the points
2 3
1 1 1
3
( 50 10 10 ) , ( 50 5 10 ) , ( 50 5 10 ) . Thus the minimum of f is
3 3
1 1 1 1
f
3
( 50 10 3 ) , ( 50+5 10 ) , ( 50+5 10 ) =
3 3 27
( 87,500 2500 10 ) , and its
1 1 1 1
maximum is f
3
( 50+10 10 ) , ( 50 5 10 ) , ( 50 5 10 ) =
3 3 27
( 87,500+2500 10 ) .
Note: If either y or z is the distinct side, then symmetry gives the same result.

2 2 2
39. We need to find the extreme values of f (x,y,z)=x +y +z subject to the two constraints
2 2
g(x,y,z)=x+y+2z=2 and h(x,y,z)=x +y  z=0 .  f = 2x,2y,2z ,   g=  , ,2 and
  h= 2 x,2 y,  . Thus we need (1) 2x= +2 x , (2) 2y= +2 y , (3) 2z=2   , (4) x+y+2z=2 ,
2 2
and (5) x +y  z=0 . From (1) and (2), 2(x y)=2 (x y) , so if x y ,  =1 . Putting this in (3) gives
1 1 1
2z=2  1 or  =z+ , but putting  =1 into (1) says  =0 . Hence z+ =0 or z= . Then (4) and (5)
2 2 2
9
Stewart Calculus ET 5e 0534393217;14. Partial Derivatives; 14.8 Lagrange Multipliers

2 2 1
become x+y 3=0 and x +y + =0 . The last equation cannot be true, so this case gives no solution.
2
2 2
So we must have x=y . Then (4) and (5) become 2x+2z=2 and 2x  z=0 which imply z=1 x and z=2x
2 2 1
. Thus 2x =1 x or 2x +x 1=(2x 1)(x+1)=0 so x= or x= 1 . The two points to check are
2
1 1 1 1 1 1 3 1 1 1
, , and (  1, 1,2 ) : f , , = and f ( 1, 1,2)=6 . Thus , , is the point
2 2 2 2 2 2 4 2 2 2
on the ellipse nearest the origin and (  1, 1,2 ) is the one farthest from the origin.

40. (a) Parametric equations for the ellipse are easiest to determine using cylindrical coordinates. The
cone is given by z=r , and the plane is 4rcos   3rsin  +8z=5 . Substituting z=r into the plane
5
equation gives 4rcos   3rsin  +8r=5 r= . Since z=r on the ellipse, parametric
4cos   3sin  +8
5
equations (in cylindrical coordinates) are  =t , r=z= , 0
t
2 .
4cos t 3sin t+8

(b) We need to find the extreme values of f (x,y,z)=z subject to the two constraints
2 2 2
g(x,y,z)=4x 3y+8z=5 and h(x,y,z)=x +y  z =0 .  f =  g+  h
2 3
0,0,1 = 4, 3,8 + 2x,2y, 2z , so we need (1) 4 +2 x=0 x= , (2)  3 +2 y=0 y= ,
 2
8  1 2 2 2
(3) 8  2 z=1 z= , (4) 4x 3y+8z=5 , and (5) x +y =z . Substituting (1), (2), and (3) into (4)
2
2 3 8  1 39  8
gives 4   3 +8 =5  = and into (5) gives
 2 2 10
2 2 3 2 8  1 2 2 2 2 2
 + =  16 +9 =(8  1)  39  16 +1=0
 2 2
1 1 1 1 4 3 5 1 1 4
 = or  = . If  = then  = and x= , y= , z= . If  = then  = and x= ,
13 3 13 2 13 13 13 3 2 3
5 4 5
y=1 , z= . Thus the highest point on the ellipse is  ,1, and the lowest point is
3 3 3
10
Stewart Calculus ET 5e 0534393217;14. Partial Derivatives; 14.8 Lagrange Multipliers

4 3 5
, , .
13 13 13

x z 2 2 2
41. f (x,y,z)=ye , g(x,y,z)=9x +4y +36z =36 , h(x,y,z)=xy+yz=1 .
x z x z x z x z
 f =  g+  h ye ,e , ye = 18x,8y,72z + y,x+z,y , so ye =18 x+ y ,
x z x z 2 2 2
e =8 y+ (x+z) ,  ye =72 z+ y , 9x +4y +36z =36 , xy+yz=1 . Using a CAS to solve these 5
equations simultaneously for x , y , z ,  , and  (in Maple, use the allvalues command), we get 4 real
 valued solutions:

x 0.222444, y  2.157012, z  0.686049,    0.200401,  2.108584

x  1.951921, y  0.545867, z 0.119973,   0.003141,    0.076238
x 0.155142, y 0.904622, z 0.950293,   0.012447,   0.489938
x 1.138731, y 1.768057, z  0.573138,   0.317141,   1.862675

Substituting these values into f gives f (0.222444, 2.157012, 0.686049)  5.3506 ,

f ( 1.951921, 0.545867,0.119973)  0.0688 , f (0.155142,0.904622,0.950293) 0.4084 ,
f (1.138731,1.768057, 0.573138) 9.7938 . Thus the maximum is approximately 9.7938 , and the
mininum is approximately  5.3506 .

2 2 2 2
42. f (x,y,z)=x+y+z , g(x,y,z)=x  y  z=0 , h(x,y,z)=x +z =4 .
 f =  g+  h 1,1,1 = 2x, 2y, 1 + 2x,0,2z , so 1=2 x+2 x , 1= 2 y , 1=  +2 z ,
2 2 2 2
x  y =z , x +z =4 . Using a CAS to solve these 5 equations simultaneously for x , y , z ,  , and  ,
we get 4 real valued solutions:

x  1.652878, y  1.964194, z  1.126052,   0.254557,    0.557060

x  1.502800, y 0.968872, z 1.319694,    0.516064,   0.183352
x  0.992513, y 1.649677, z  1.736352,    0.303090,    0.200682
x 1.895178, y 1.718347, z 0.638984,    0.290977,   0.554805

Substituting these values into f gives f ( 1.652878, 1.964194, 1.126052)  4.7431 ,

f ( 1.502800,0.968872,1.319694) 0.7858 , f ( 0.992513,1.649677, 1.736352)  1.0792 ,
f (1.895178,1.718347,0.638984) 4.2525 . Thus the maximum is approximately 4.2525 , and the
mininum is approximately  4.7431 .

43. (a) We wish to maximize f x ,x ,... ,x = n x x    x subject to

( )
1 2 n 1 2 n
g x ,x ,... ,x =x +x +   +x =c and x >0 . 
( ) f
1 2 n 1 2 n i

11
Stewart Calculus ET 5e 0534393217;14. Partial Derivatives; 14.8 Lagrange Multipliers

1 1 1
 1  1  1
1 n 1 n 1 n
= x x    x
( ) ( x    x ) , n ( x x    x ) ( x x    x ) ,... , n ( x x    x ) ( x    x )
n 1 2 n 2 n 1 2 n 1 3 n 1 2 n 1 n 1

and
  g=  , ,... , , so we need to solve the system of equations
1
 1 1/n 1/n 1/n
1 n  x x    x =n x
x x    x
( ) ( x    x ) = 1 2 n 1
n 1 2 n 2 n
1
 1 1/n 1/n 1/n
1 n  x x    x =n x
x x    x
( ) ( x x    x ) = 1 2 n 2
n 1 2 n 1 3 n
  
  
  
1
 1 1/n 1/n 1/n
1 n  x x    x =n x
x x    x
( x    x ) (= 1 2 n n
)
n 1 2 n 1 n 1

This implies n x =n x =   =n x . Note   0 , otherwise we can’t have all x >0 . Thus
1 2 n i
c
x =x =   =x . But x +x +   +x =c nx =c  x= =x =x =   =x . Then the only point where
1 2 n 1 2 n 1 1 n 2 3 n
c c c
f can have an extreme value is , ,... ,
n n n
. Since we can choose values for x ,x ,... ,x that
1 2 n ( )
make f as close to zero (but not equal) as we like, f has no minimum value. Thus the maximum
c c c n c c c c
value is f , ,... , =       = .
n n n n n n n
c c
(b) From part (a), is the maximum value of f . Thus f x ,x ,... ,x = n x x    x
. But ( )
n 1 2 n 1 2 n n
x +x +   +x
1 2 n
x +x +   +x =c , so x x    x

n
. These two means are equal when f attains its
1 2 n 1 2 n n
c c c c
maximum value , but this can occur only at the point , ,... , we found in part (a). So
n n n n
c
the means are equal only when x =x =x =   =x = .
1 2 3 n n

44. (a) Let f x ,... ,x ,y ,... ,y = x y , g x ,... ,x = x , and h x ,... ,x =

n n 2 n 2
( 1 n 1 n ) i =1 i i ( 1 n ) i =1 i ( 1 n ) y . Then
i =1 i

 
n n 2
 f = x y = y ,y ,... ,y ,x ,x ,... ,x ,  g= x = 2x ,2x ,... ,2x ,0,0,... ,0 and
i =1 i i 1 2 n 1 2 n i =1 i 1 2 n
12
Stewart Calculus ET 5e 0534393217;14. Partial Derivatives; 14.8 Lagrange Multipliers


n 2
 h= y = 0,0,... ,0,2y ,2y ,... ,2y . So  f =  g+  h y =2 x and x =2 y , 1
i
n .
i =1 i 1 2 n i i i i
1
Then 1= y = 
n 2 n 2 2 2 n 2 2
4 x =4 x =4   = .
i =1 i i =1 i i =1 i 2
1 1 1
x =x , 1
i
n . Thus  x y =
n n 2
If  = then y =2 x =1 . Similarly if  = we get
2 i 2 i i i =1 i i i =1 i 2
1
y = x and  giving y = x , 1
i
n , and 
n n
x y = 1 . Similarly we get  = x y = 1 . Thus
i i i =1 i i 2 i i i =1 i i

the maximum value of 

n
x y is 1 .
i =1 i i

(b) Here we assume  a  0 and  b  0 . (If 

n 2 n 2 n 2
a =0 , then each a =0 and so the inequality
i =1 i i =1 i i =1 i i

 
2 2
a a b b
 
i 2 i i 2 i
is trivially true.) x =  x= =1 , and y =  y= =1 . Therefore, from
i i i i
 
2 2
 
2 2
a a b b
j j j j

ab
(a),  x y =   
i i 2 2

1 a b
a b .
i i i i j j
 
2 2
a b
j j

13