From my research, this is what i got

VGG16:

• Consists of 16 layers (13 convolutional and 3 fully connected layers).
• Uses 3×3 convolutional layers stacked on top of each other in increasing depth.
• Reduces volume size by using max pooling.
• Connected to three fully-connected layers, which then connect to a softmax classifier1.

VGG19:

• Similar to VGG16 but with 19 layers (16 convolutional and 3 fully connected layers).

VGG19 has more layers, does that mean it’s better?